Solubility is a property of central importance for the use of proteins in research in molecular and cell biology and in applications in biotechnology and medicine. Since experimental methods for measuring protein solubility are material intensive and time consuming, computational methods have recently emerged to enable the rapid and inexpensive screening of solubility for large libraries of proteins, as it is routinely required in development pipelines. Here, we describe the development of one such method to include in the predictions the effect of the pH on solubility. We illustrate the resulting pH-dependent predictions on a variety of antibodies and other proteins to demonstrate that these predictions achieve an accuracy comparable with that of experimental methods. We make this method publicly available at https://www-cohsoftware.ch.cam.ac.uk/index.php/camsolph, as the version 3.0 of CamSol.

Recently, extracting inherent biological system information (e.g. cellular networks) from genome-wide expression profiles for developing personalized diagnostic and therapeutic strategies has become increasingly important. However, accurately constructing single-sample networks (SINs) to capture individual characteristics and heterogeneity in disease remains challenging. Here, we propose a sample-specific-weighted correlation network (SWEET) method to model SINs by integrating the genome-wide sample-to-sample correlation (i.e. sample weights) with the differential network between perturbed and aggregate networks. For a group of samples, the genome-wide sample weights can be assessed without prior knowledge of intrinsic subpopulations to address the network edge number bias caused by sample size differences. Compared with the state-of-the-art SIN inference methods, the SWEET SINs in 16 cancers more likely fit the scale-free property, display higher overlap with the human interactomes and perform better in identifying three types of cancer-related genes. Moreover, integrating SWEET SINs with a network proximity measure facilitates characterizing individual features and therapy in diseases, such as somatic mutation, mut-driver and essential genes. Biological experiments further validated two candidate repurposable drugs, albendazole for head and neck squamous cell carcinoma (HNSCC) and lung adenocarcinoma (LUAD) and encorafenib for HNSCC. By applying SWEET, we also identified two possible LUAD subtypes that exhibit distinct clinical features and molecular mechanisms. Overall, the SWEET method complements current SIN inference and analysis methods and presents a view of biological systems at the network level to offer numerous clues for further investigation and clinical translation in network medicine and precision medicine.

With the emergence of high-throughput technologies, computational screening based on gene expression profiles has become one of the most effective methods for drug discovery. More importantly, profile-based approaches remarkably enhance novel drug-disease pair discovery without relying on drug- or disease-specific prior knowledge, which has been widely used in modern medicine. However, profile-based systematic screening of active ingredients of traditional Chinese medicine (TCM) has been scarcely performed due to inadequate pharmacotranscriptomic data. Here, we develop the largest-to-date online TCM active ingredients-based pharmacotranscriptomic platform integrated traditional Chinese medicine (ITCM) for the effective screening of active ingredients. First, we performed unified high-throughput experiments and constructed the largest data repository of 496 representative active ingredients, which was five times larger than the previous one built by our team. The transcriptome-based multi-scale analysis was also performed to elucidate their mechanism. Then, we developed six state-of-art signature search methods to screen active ingredients and determine the optimal signature size for all methods. Moreover, we integrated them into a screening strategy, TCM-Query, to identify the potential active ingredients for the special disease. In addition, we also comprehensively collected the TCM-related resource by literature mining. Finally, we applied ITCM to an active ingredient bavachinin, and two diseases, including prostate cancer and COVID-19, to demonstrate the power of drug discovery. ITCM was aimed to comprehensively explore the active ingredients of TCM and boost studies of pharmacological action and drug discovery. ITCM is available at http://itcm.biotcm.net.

MOTIVATION: Predicting the associations between human microbes and drugs (MDAs) is one critical step in drug development and precision medicine areas. Since discovering these associations through wet experiments is time-consuming and labor-intensive, computational methods have already been an effective way to tackle this problem. Recently, graph contrastive learning (GCL) approaches have shown great advantages in learning the embeddings of nodes from heterogeneous biological graphs (HBGs). However, most GCL-based approaches don't fully capture the rich structure information in HBGs. Besides, fewer MDA prediction methods could screen out the most informative negative samples for effectively training the classifier. Therefore, it still needs to improve the accuracy of MDA predictions. RESULTS: In this study, we propose a novel approach that employs the Structure-enhanced Contrastive learning and Self-paced negative sampling strategy for Microbe-Drug Association predictions (SCSMDA). Firstly, SCSMDA constructs the similarity networks of microbes and drugs, as well as their different meta-path-induced networks. Then SCSMDA employs the representations of microbes and drugs learned from meta-path-induced networks to enhance their embeddings learned from the similarity networks by the contrastive learning strategy. After that, we adopt the self-paced negative sampling strategy to select the most informative negative samples to train the MLP classifier. Lastly, SCSMDA predicts the potential microbe-drug associations with the trained MLP classifier. The embeddings of microbes and drugs learning from the similarity networks are enhanced with the contrastive learning strategy, which could obtain their discriminative representations. Extensive results on three public datasets indicate that SCSMDA significantly outperforms other baseline methods on the MDA prediction task. Case studies for two common drugs could further demonstrate the effectiveness of SCSMDA in finding novel MDA associations. AVAILABILITY: The source code is publicly available on GitHub https://github.com/Yue-Yuu/SCSMDA-master.

N6-methyladinosine (m6A) modification is the most abundant co-transcriptional modification in eukaryotic RNA and plays important roles in cellular regulation. Traditional high-throughput sequencing experiments used to explore functional mechanisms are time-consuming and labor-intensive, and most of the proposed methods focused on limited species types. To further understand the relevant biological mechanisms among different species with the same RNA modification, it is necessary to develop a computational scheme that can be applied to different species. To achieve this, we proposed an attention-based deep learning method, adaptive-m6A, which consists of convolutional neural network, bi-directional long short-term memory and an attention mechanism, to identify m6A sites in multiple species. In addition, three conventional machine learning (ML) methods, including support vector machine, random forest and logistic regression classifiers, were considered in this work. In addition to the performance of ML methods for multi-species prediction, the optimal performance of adaptive-m6A yielded an accuracy of 0.9832 and the area under the receiver operating characteristic curve of 0.98. Moreover, the motif analysis and cross-validation among different species were conducted to test the robustness of one model towards multiple species, which helped improve our understanding about the sequence characteristics and biological functions of RNA modifications in different species.

The advance in single-cell RNA-sequencing (scRNA-seq) sheds light on cell-specific transcriptomic studies of cell developments, complex diseases and cancers. Nevertheless, scRNA-seq techniques suffer from 'dropout' events, and imputation tools are proposed to address the sparsity. Here, rather than imputation, we propose a tool, SMURF, to extract the low-dimensional embeddings from cells and genes utilizing matrix factorization with a mixture of Poisson-Gamma divergent as objective while preserving self-consistency. SMURF exhibits feasible cell subpopulation discovery efficacy with obtained cell embeddings on replicated in silico and eight web lab scRNA datasets with ground truth cell types. Furthermore, SMURF can reduce the cell embedding to a 1D-oval space to recover the time course of cell cycle. SMURF can also serve as an imputation tool; the in silico data assessment shows that SMURF parades the most robust gene expression recovery power with low root mean square error and high Pearson correlation. Moreover, SMURF recovers the gene distribution for the WM989 Drop-seq data. SMURF is available at https://github.com/deepomicslab/SMURF.

A large number of works have presented the single-cell RNA sequencing (scRNA-seq) to study the diversity and biological functions of cells at the single-cell level. Clustering identifies unknown cell types, which is essential for downstream analysis of scRNA-seq samples. However, the high dimensionality, high noise and pervasive dropout rate of scRNA-seq samples have a significant challenge to the cluster analysis of scRNA-seq samples. Herein, we propose a new adaptive fuzzy clustering model based on the denoising autoencoder and self-attention mechanism called the scDASFK. It implements the comparative learning to integrate cell similar information into the clustering method and uses a deep denoising network module to denoise the data. scDASFK consists of a self-attention mechanism for further denoising where an adaptive clustering optimization function for iterative clustering is implemented. In order to make the denoised latent features better reflect the cell structure, we introduce a new adaptive feedback mechanism to supervise the denoising process through the clustering results. Experiments on 16 real scRNA-seq datasets show that scDASFK performs well in terms of clustering accuracy, scalability and stability. Overall, scDASFK is an effective clustering model with great potential for scRNA-seq samples analysis. Our scDASFK model codes are freely available at https://github.com/LRX2022/scDASFK.

Predicting therapeutic responses in cancer patients is a major challenge in the field of precision medicine due to high inter- and intra-tumor heterogeneity. Most drug response models need to be improved in terms of accuracy, and there is limited research to assess therapeutic responses of particular tumor types. Here, we developed a novel method DROEG (Drug Response based on Omics and Essential Genes) for prediction of drug response in tumor cell lines by integrating genomic, transcriptomic and methylomic data along with CRISPR essential genes, and revealed that the incorporation of tumor proliferation essential genes can improve drug sensitivity prediction. Concisely, DROEG integrates literature-based and statistics-based methods to select features and uses Support Vector Regression for model construction. We demonstrate that DROEG outperforms most state-of-the-art algorithms by both qualitative (prediction accuracy for drug-sensitive/resistant) and quantitative (Pearson correlation coefficient between the predicted and actual IC50) evaluation in Genomics of Drug Sensitivity in Cancer and Cancer Cell Line Encyclopedia datasets. In addition, DROEG is further applied to the pan-gastrointestinal tumor with high prevalence and mortality as a case study at both cell line and clinical levels to evaluate the model efficacy and discover potential prognostic biomarkers in Cisplatin and Epirubicin treatment. Interestingly, the CRISPR essential gene information is found to be the most important contributor to enhance the accuracy of the DROEG model. To our knowledge, this is the first study to integrate essential genes with multi-omics data to improve cancer drug response prediction and provide insights into personalized precision treatment.

Complex biological systems do not always develop smoothly but occasionally undergo a sharp transition; i.e. there exists a critical transition or tipping point at which a drastic qualitative shift occurs. Hunting for such a critical transition is important to prevent or delay the occurrence of catastrophic consequences, such as disease deterioration. However, the identification of the critical state for complex biological systems is still a challenging problem when using high-dimensional small sample data, especially where only a certain sample is available, which often leads to the failure of most traditional statistical approaches. In this study, a novel quantitative method, sample-perturbed network entropy (SPNE), is developed based on the sample-perturbed directed network to reveal the critical state of complex biological systems at the single-sample level. Specifically, the SPNE approach effectively quantifies the perturbation effect caused by a specific sample on the directed network in terms of network entropy and thus captures the criticality of biological systems. This model-free method was applied to both bulk and single-cell expression data. Our approach was validated by successfully detecting the early warning signals of the critical states for six real datasets, including four tumor datasets from The Cancer Genome Atlas (TCGA) and two single-cell datasets of cell differentiation. In addition, the functional analyses of signaling biomarkers demonstrated the effectiveness of the analytical and computational results.

Microbes can affect the metabolism and immunity of human body incessantly, and the dysbiosis of human microbiome drives not only the occurrence but also the progression of disease (i.e. multiple statuses of disease). Recently, microbiome-based association tests have been widely developed to detect the association between the microbiome and host phenotype. However, the existing methods have not achieved satisfactory performance in testing the association between the microbiome and ordinal/nominal multicategory phenotypes (e.g. disease severity and tumor subtype). In this paper, we propose an optimal microbiome-based association test for multicategory phenotypes, namely, multiMiAT. Specifically, under the multinomial logit model framework, we first introduce a microbiome regression-based kernel association test for multicategory phenotypes (multiMiRKAT). As a data-driven optimal test, multiMiAT then integrates multiMiRKAT, score test and MiRKAT-MC to maintain excellent performance in diverse association patterns. Massive simulation experiments prove the success of our method. Furthermore, multiMiAT is also applied to real microbiome data experiments to detect the association between the gut microbiome and clinical statuses of colorectal cancer as well as for diverse statuses of Clostridium difficile infections.

Due to the high heterogeneity and complexity of cancers, patients with different cancer subtypes often have distinct groups of genomic and clinical characteristics. Therefore, the discovery and identification of cancer subtypes are crucial to cancer diagnosis, prognosis and treatment. Recent technological advances have accelerated the increasing availability of multi-omics data for cancer subtyping. To take advantage of the complementary information from multi-omics data, it is necessary to develop computational models that can represent and integrate different layers of data into a single framework. Here, we propose a decoupled contrastive clustering method (Subtype-DCC) based on multi-omics data integration for clustering to identify cancer subtypes. The idea of contrastive learning is introduced into deep clustering based on deep neural networks to learn clustering-friendly representations. Experimental results demonstrate the superior performance of the proposed Subtype-DCC model in identifying cancer subtypes over the currently available state-of-the-art clustering methods. The strength of Subtype-DCC is also supported by the survival and clinical analysis.

Recognizing binding sites of DNA-binding proteins is a key factor for elucidating transcriptional regulation in organisms. ChIP-exo enables researchers to delineate genome-wide binding landscapes of DNA-binding proteins with near single base-pair resolution. However, the peak calling step hinders ChIP-exo application since the published algorithms tend to generate false-positive and false-negative predictions. Here, we report the development of DEOCSU (DEep-learning Optimized ChIP-exo peak calling SUite), a novel machine learning-based ChIP-exo peak calling suite. DEOCSU entails the deep convolutional neural network model which was trained with curated ChIP-exo peak data to distinguish the visualized data of bona fide peaks from false ones. Performance validation of the trained deep-learning model indicated its high accuracy, high precision and high recall of over 95%. Applying the new suite to both in-house and publicly available ChIP-exo datasets obtained from bacteria, eukaryotes and archaea revealed an accurate prediction of peaks containing canonical motifs, highlighting the versatility and efficiency of DEOCSU. Furthermore, DEOCSU can be executed on a cloud computing platform or the local environment. With visualization software included in the suite, adjustable options such as the threshold of peak probability, and iterable updating of the pre-trained model, DEOCSU can be optimized for users' specific needs.

Protein arginine methylation is an important posttranslational modification (PTM) associated with protein functional diversity and pathological conditions including cancer. Identification of methylation binding sites facilitates a better understanding of the molecular function of proteins. Recent developments in the field of deep neural networks have led to a proliferation of deep learning-based methylation identification studies because of their fast and accurate prediction. In this paper, we propose DeepGpgs, an advanced deep learning model incorporating Gaussian prior and gated attention mechanism. We introduce a residual network channel to extract the evolutionary information of proteins. Then we combine the adaptive embedding with bidirectional long short-term memory networks to form a context-shared encoder layer. A gated multi-head attention mechanism is followed to obtain the global information about the sequence. A Gaussian prior is injected into the sequence to assist in predicting PTMs. We also propose a weighted joint loss function to alleviate the false negative problem. We empirically show that DeepGpgs improves Matthews correlation coefficient by 6.3% on the arginine methylation independent test set compared with the existing state-of-the-art methylation site prediction methods. Furthermore, DeepGpgs has good robustness in phosphorylation site prediction of SARS-CoV-2, which indicates that DeepGpgs has good transferability and the potential to be extended to other modification sites prediction. The open-source code and data of the DeepGpgs can be obtained from https://github.com/saizhou1/DeepGpgs.

Cryo-electron microscopy (cryo-EM) allows a macromolecular structure such as protein-DNA/RNA complexes to be reconstructed in a three-dimensional coulomb potential map. The structural information of these macromolecular complexes forms the foundation for understanding the molecular mechanism including many human diseases. However, the model building of large macromolecular complexes is often difficult and time-consuming. We recently developed DeepTracer-2.0, an artificial-intelligence-based pipeline that can build amino acid and nucleic acid backbones from a single cryo-EM map, and even predict the best-fitting residues according to the density of side chains. The experiments showed improved accuracy and efficiency when benchmarking the performance on independent experimental maps of protein-DNA/RNA complexes and demonstrated the promising future of macromolecular modeling from cryo-EM maps. Our method and pipeline could benefit researchers worldwide who work in molecular biomedicine and drug discovery, and substantially increase the throughput of the cryo-EM model building. The pipeline has been integrated into the web portal https://deeptracer.uw.edu/.

During intracellular signal transduction, protein-protein interactions (PPIs) facilitate protein complex assembly to regulate protein localization and function, which are critical for numerous cellular events. Over the years, multiple techniques have been developed to characterize PPIs to elucidate roles and regulatory mechanisms of proteins. Among them, the mass spectrometry (MS)-based interactome analysis has been increasing in popularity due to its unbiased and informative manner towards understanding PPI networks. However, with MS instrumentation advancing and yielding more data than ever, the analysis of a large amount of PPI-associated proteomic data to reveal bona fide interacting proteins become challenging. Here, we review the methods and bioinformatic resources that are commonly used in analyzing large interactome-related proteomic data and propose a simple guideline for identifying novel interacting proteins for biological research.

Due to the lack of a method to efficiently represent the multimodal information of a protein, including its structure and sequence information, predicting compound-protein binding affinity (CPA) still suffers from low accuracy when applying machine-learning methods. To overcome this limitation, in a novel end-to-end architecture (named FeatNN), we develop a coevolutionary strategy to jointly represent the structure and sequence features of proteins and ultimately optimize the mathematical models for predicting CPA. Furthermore, from the perspective of data-driven approach, we proposed a rational method that can utilize both high- and low-quality databases to optimize the accuracy and generalization ability of FeatNN in CPA prediction tasks. Notably, we visually interpret the feature interaction process between sequence and structure in the rationally designed architecture. As a result, FeatNN considerably outperforms the state-of-the-art (SOTA) baseline in virtual drug evaluation tasks, indicating the feasibility of this approach for practical use. FeatNN provides an outstanding method for higher CPA prediction accuracy and better generalization ability by efficiently representing multimodal information of proteins via a coevolutionary strategy.

A-to-I RNA editing diversifies human transcriptome to confer its functional effects on the downstream genes or regulations, potentially involving in neurodegenerative pathogenesis. Its variabilities are attributed to multiple regulators, including the key factor of genetic variants. To comprehensively investigate the potentials of neurodegenerative disease-susceptibility variants from the view of A-to-I RNA editing, we analyzed matched genetic and transcriptomic data of 1596 samples across nine brain tissues and whole blood from two large consortiums, Accelerating Medicines Partnership-Alzheimer's Disease and Parkinson's Progression Markers Initiative. The large-scale and genome-wide identification of 95 198 RNA editing quantitative trait loci revealed the preferred genetic effects on adjacent editing events. Furthermore, to explore the underlying mechanisms of the genetic controls of A-to-I RNA editing, several top RNA-binding proteins were pointed out, such as EIF4A3, U2AF2, NOP58, FBL, NOP56 and DHX9, since their regulations on multiple RNA-editing events were probably interfered by these genetic variants. Moreover, these variants may also contribute to the variability of other molecular phenotypes associated with RNA editing, including the functions of 3 proteins, expressions of 277 genes and splicing of 449 events. All the analyses results shown in NeuroEdQTL (https://relab.xidian.edu.cn/NeuroEdQTL/) constituted a unique resource for the understanding of neurodegenerative pathogenesis from genotypes to phenotypes related to A-to-I RNA editing.

Erratum in Brief Bioinform. 2023 Feb 01;:

Somatic mutational signatures (MSs) identified by genome sequencing play important roles in exploring the cause and development of cancer. Thus far, many such signatures have been identified, and some of them do imply causes of cancer. However, a major bottleneck is that we do not know the potential meanings (i.e. carcinogenesis or biological functions) and contributing genes for most of them. Here, we presented a computational framework, Gene Somatic Genome Pattern (GSGP), which can decipher the molecular mechanisms of the MSs. More importantly, it is the first time that the GSGP is able to process MSs from ribonucleic acid (RNA) sequencing, which greatly extended the applications of both MS analysis and RNA sequencing (RNAseq). As a result, GSGP analyses match consistently with previous reports and identify the etiologies for a number of novel signatures. Notably, we applied GSGP to RNAseq data and revealed an RNA-derived MS involved in deficient deoxyribonucleic acid mismatch repair and microsatellite instability in colorectal cancer. Researchers can perform customized GSGP analysis using the web tools or scripts we provide.

While deep learning (DL)-based models have emerged as powerful approaches to predict protein-protein interactions (PPIs), the reliance on explicit similarity measures (e.g. sequence similarity and network neighborhood) to known interacting proteins makes these methods ineffective in dealing with novel proteins. The advent of AlphaFold2 presents a significant opportunity and also a challenge to predict PPIs in a straightforward way based on monomer structures while controlling bias from protein sequences. In this work, we established Structure and Graph-based Predictions of Protein Interactions (SGPPI), a structure-based DL framework for predicting PPIs, using the graph convolutional network. In particular, SGPPI focused on protein patches on the protein-protein binding interfaces and extracted the structural, geometric and evolutionary features from the residue contact map to predict PPIs. We demonstrated that our model outperforms traditional machine learning methods and state-of-the-art DL-based methods using non-representation-bias benchmark datasets. Moreover, our model trained on human dataset can be reliably transferred to predict yeast PPIs, indicating that SGPPI can capture converging structural features of protein interactions across various species. The implementation of SGPPI is available at https://github.com/emerson106/SGPPI.

Single-cell RNA-seq enabled in-depth study on tissue micro-environment and immune-profiling, where a crucial step is to annotate cell identity. Immune cells play key roles in many diseases, whereas their activities are hard to track due to their diverse and highly variable nature. Existing cell-type identifiers had limited performance for this purpose. We present HiCAT, a hierarchical, marker-based cell-type identifier utilising gene set analysis for statistical scoring for given markers. It features successive identification of major-type, minor-type and subsets utilising subset markers structured in a three-level taxonomy tree. Comparison with manual annotation and pairwise match test showed HiCAT outperforms others in major- and minor-type identification. For subsets, we qualitatively evaluated the marker expression profile demonstrating that HiCAT provide the clearest immune-cell landscape. HiCAT was also used for immune-cell profiling in ulcerative colitis and discovered distinct features of the disease in macrophage and T-cell subsets that could not be identified previously.

Binding affinity prediction largely determines the discovery efficiency of lead compounds in drug discovery. Recently, machine learning (ML)-based approaches have attracted much attention in hopes of enhancing the predictive performance of traditional physics-based approaches. In this study, we evaluated the impact of structural dynamic information on the binding affinity prediction by comparing the models trained on different dimensional descriptors, using three targets (i.e. JAK1, TAF1-BD2 and DDR1) and their corresponding ligands as the examples. Here, 2D descriptors are traditional ECFP4 fingerprints, 3D descriptors are the energy terms of the Smina and NNscore scoring functions and 4D descriptors contain the structural dynamic information derived from the trajectories based on molecular dynamics (MD) simulations. We systematically investigate the MD-refined binding affinity prediction performance of three classical ML algorithms (i.e. RF, SVR and XGB) as well as two common virtual screening methods, namely Glide docking and MM/PBSA. The outcomes of the ML models built using various dimensional descriptors and their combinations reveal that the MD refinement with the optimized protocol can improve the predictive performance on the TAF1-BD2 target with considerable structural flexibility, but not for the less flexible JAK1 and DDR1 targets, when taking docking poses as the initial structure instead of the crystal structures. The results highlight the importance of the initial structures to the final performance of the model through conformational analysis on the three targets with different flexibility.

The determination of transcriptome profiles that mediate immune therapy in cancer remains a major clinical and biological challenge. Despite responses induced by immune-check points inhibitors (ICIs) in diverse tumor types and all the big breakthroughs in cancer immunotherapy, most patients with solid tumors do not respond to ICI therapies. It still remains a big challenge to predict the ICI treatment response. Here, we propose a framework with multiple prior knowledge networks guided for immune checkpoints inhibitors prediction-DeepOmix-ICI (or ICInet for short). ICInet can predict the immune therapy response by leveraging geometric deep learning and prior biological knowledge graphs of gene-gene interactions. Here, we demonstrate more than 600 ICI-treated patients with ICI response data and gene expression profile to apply on ICInet. ICInet was used for ICI therapy responses prediciton across different cancer types-melanoma, gastric cancer and bladder cancer, which includes 7 cohorts from different data sources. ICInet is able to robustly generalize into multiple cancer types. Moreover, the performance of ICInet in those cancer types can outperform other ICI biomarkers in the clinic. Our model [area under the curve (AUC = 0.85)] generally outperformed other measures, including tumor mutational burden (AUC = 0.62) and programmed cell death ligand-1 score (AUC = 0.74). Therefore, our study presents a prior-knowledge guided deep learning method to effectively select immunotherapy-response-associated biomarkers, thereby improving the prediction of immunotherapy response for precision oncology.

Identification of potential targets for known bioactive compounds and novel synthetic analogs is of considerable significance. In silico target fishing (TF) has become an alternative strategy because of the expensive and laborious wet-lab experiments, explosive growth of bioactivity data and rapid development of high-throughput technologies. However, these TF methods are based on different algorithms, molecular representations and training datasets, which may lead to different results when predicting the same query molecules. This can be confusing for practitioners in practical applications. Therefore, this study systematically evaluated nine popular ligand-based TF methods based on target and ligand-target pair statistical strategies, which will help practitioners make choices among multiple TF methods. The evaluation results showed that SwissTargetPrediction was the best method to produce the most reliable predictions while enriching more targets. High-recall similarity ensemble approach (SEA) was able to find real targets for more compounds compared with other TF methods. Therefore, SwissTargetPrediction and SEA can be considered as primary selection methods in future studies. In addition, the results showed that k = 5 was the optimal number of experimental candidate targets. Finally, a novel ensemble TF method based on consensus voting is proposed to improve the prediction performance. The precision of the ensemble TF method outperforms the individual TF method, indicating that the ensemble TF method can more effectively identify real targets within a given top-k threshold. The results of this study can be used as a reference to guide practitioners in selecting the most effective methods in computational drug discovery.

Extensive investigation of gene fusions in cancer has led to the discovery of novel biomarkers and therapeutic targets. To date, most studies have neglected chromosomal rearrangement-independent fusion transcripts and complex fusion structures such as double or triple-hop fusions, and fusion-circRNAs. In this review, we untangle fusion-related terminology and propose a classification system involving both gene and transcript fusions. We highlight the importance of RNA-level fusions and how long-read sequencing approaches can improve detection and characterization. Moreover, we discuss novel bioinformatic tools to identify fusions in long-read sequencing data and strategies to experimentally validate and functionally characterize fusion transcripts.

The label-free quantification (LFQ) has emerged as an exceptional technique in proteomics owing to its broad proteome coverage, great dynamic ranges and enhanced analytical reproducibility. Due to the extreme difficulty lying in an in-depth quantification, the LFQ chains incorporating a variety of transformation, pretreatment and imputation methods are required and constructed. However, it remains challenging to determine the well-performing chain, owing to its strong dependence on the studied data and the diverse possibility of integrated chains. In this study, an R package EVALFQ was therefore constructed to enable a performance evaluation on >3000 LFQ chains. This package is unique in (a) automatically evaluating the performance using multiple criteria, (b) exploring the quantification accuracy based on spiking proteins and (c) discovering the well-performing chains by comprehensive assessment. All in all, because of its superiority in assessing from multiple perspectives and scanning among over 3000 chains, this package is expected to attract broad interests from the fields of proteomic quantification. The package is available at https://github.com/idrblab/EVALFQ.

The analysis of the combined mRNA and miRNA content of a biological sample can be of interest for answering several research questions, like biomarkers discovery, or mRNA-miRNA interactions. However, the process is costly and time-consuming, separate libraries need to be prepared and sequenced on different flowcells. Combo-Seq is a library prep kit that allows us to prepare combined mRNA-miRNA libraries starting from very low total RNA. To date, no dedicated bioinformatics method exists for the processing of Combo-Seq data. In this paper, we describe CODA (Combo-seq Data Analysis), a workflow specifically developed for the processing of Combo-Seq data that employs existing free-to-use tools. We compare CODA with exceRpt, the pipeline suggested by the kit manufacturer for this purpose. We also evaluate how Combo-Seq libraries analysed with CODA perform compared with conventional poly(A) and small RNA libraries prepared from the same samples. We show that using CODA more successfully trimmed reads are recovered compared with exceRpt, and the difference is more dramatic with short sequencing reads. We demonstrate how Combo-Seq identifies as many genes and fewer miRNAs compared to the standard libraries, and how miRNA validation favours conventional small RNA libraries over Combo-Seq. The CODA code is available at https://github.com/marta-nazzari/CODA.

Lysine succinylation is a kind of post-translational modification (PTM) that plays a crucial role in regulating the cellular processes. Aberrant succinylation may cause inflammation, cancers, metabolism diseases and nervous system diseases. The experimental methods to detect succinylation sites are time-consuming and costly. This thus calls for computational models with high efficacy, and attention has been given in the literature to develop such models, albeit with only moderate success in the context of different evaluation metrics. One crucial aspect in this context is the biochemical and physicochemical properties of amino acids, which appear to be useful as features for such computational predictors. However, some of the existing computational models did not use the biochemical and physicochemical properties of amino acids. In contrast, some others used them without considering the inter-dependency among the properties. The combinations of biochemical and physicochemical properties derived through our optimization process achieve better results than the results achieved by combining all the properties. We propose three deep learning architectures: CNN+Bi-LSTM (CBL), Bi-LSTM+CNN (BLC) and their combination (CBL_BLC). We find that CBL_BLC outperforms the other two. Ensembling of different models successfully improves the results. Notably, tuning the threshold of the ensemble classifiers further improves the results. Upon comparing our work with other existing works on two datasets, we successfully achieve better sensitivity and specificity by varying the threshold value.

Deoxyribonucleic acid(DNA) N6-methyladenine plays a vital role in various biological processes, and the accurate identification of its site can provide a more comprehensive understanding of its biological effects. There are several methods for 6mA site prediction. With the continuous development of technology, traditional techniques with the high costs and low efficiencies are gradually being replaced by computer methods. Computer methods that are widely used can be divided into two categories: traditional machine learning and deep learning methods. We first list some existing experimental methods for predicting the 6mA site, then analyze the general process from sequence input to results in computer methods and review existing model architectures. Finally, the results were summarized and compared to facilitate subsequent researchers in choosing the most suitable method for their work.

Accurate in silico prediction of conformational B-cell epitopes would lead to major improvements in disease diagnostics, drug design and vaccine development. A variety of computational methods, mainly based on machine learning approaches, have been developed in the last decades to tackle this challenging problem. Here, we rigorously benchmarked nine state-of-the-art conformational B-cell epitope prediction webservers, including generic and antibody-specific methods, on a dataset of over 250 antibody-antigen structures. The results of our assessment and statistical analyses show that all the methods achieve very low performances, and some do not perform better than randomly generated patches of surface residues. In addition, we also found that commonly used consensus strategies that combine the results from multiple webservers are at best only marginally better than random. Finally, we applied all the predictors to the SARS-CoV-2 spike protein as an independent case study, and showed that they perform poorly in general, which largely recapitulates our benchmarking conclusions. We hope that these results will lead to greater caution when using these tools until the biases and issues that limit current methods have been addressed, promote the use of state-of-the-art evaluation methodologies in future publications and suggest new strategies to improve the performance of conformational B-cell epitope prediction methods.

The coronavirus disease of 2019 pandemic has catalyzed the rapid development of mRNA vaccines, whereas, how to optimize the mRNA sequence of exogenous gene such as severe acute respiratory syndrome coronavirus 2 spike to fit human cells remains a critical challenge. A new algorithm, iDRO (integrated deep-learning-based mRNA optimization), is developed to optimize multiple components of mRNA sequences based on given amino acid sequences of target protein. Considering the biological constraints, we divided iDRO into two steps: open reading frame (ORF) optimization and 5' untranslated region (UTR) and 3'UTR generation. In ORF optimization, BiLSTM-CRF (bidirectional long-short-term memory with conditional random field) is employed to determine the codon for each amino acid. In UTR generation, RNA-Bart (bidirectional auto-regressive transformer) is proposed to output the corresponding UTR. The results show that the optimized sequences of exogenous genes acquired the pattern of human endogenous gene sequence. In experimental validation, the mRNA sequence optimized by our method, compared with conventional method, shows higher protein expression. To the best of our knowledge, this is the first study by introducing deep-learning methods to integrated mRNA sequence optimization, and these results may contribute to the development of mRNA therapeutics.

In epistasis analysis, single-nucleotide polymorphism-single-nucleotide polymorphism interactions (SSIs) among genes may, alongside other environmental factors, influence the risk of multifactorial diseases. To identify SSI between cases and controls (i.e. binary traits), the score for model quality is affected by different objective functions (i.e. measurements) because of potential disease model preferences and disease complexities. Our previous study proposed a multiobjective approach-based multifactor dimensionality reduction (MOMDR), with the results indicating that two objective functions could enhance SSI identification with weak marginal effects. However, SSI identification using MOMDR remains a challenge because the optimal measure combination of objective functions has yet to be investigated. This study extended MOMDR to the many-objective version (i.e. many-objective MDR, MaODR) by integrating various disease probability measures based on a two-way contingency table to improve the identification of SSI between cases and controls. We introduced an objective function selection approach to determine the optimal measure combination in MaODR among 10 well-known measures. In total, 6 disease models with and 40 disease models without marginal effects were used to evaluate the general algorithms, namely those based on multifactor dimensionality reduction, MOMDR and MaODR. Our results revealed that the MaODR-based three objective function model, correct classification rate, likelihood ratio and normalized mutual information (MaODR-CLN) exhibited the higher 6.47% detection success rates (Accuracy) than MOMDR and higher 17.23% detection success rates than MDR through the application of an objective function selection approach. In a Wellcome Trust Case Control Consortium, MaODR-CLN successfully identified the significant SSIs (P < 0.001) associated with coronary artery disease. We performed a systematic analysis to identify the optimal measure combination in MaODR among 10 objective functions. Our combination detected SSIs-based binary traits with weak marginal effects and thus reduced spurious variables in the score model. MOAI is freely available at https://sites.google.com/view/maodr/home.

Viruses are the most ubiquitous and diverse entities in the biome. Due to the rapid growth of newly identified viruses, there is an urgent need for accurate and comprehensive virus classification, particularly for novel viruses. Here, we present PhaGCN2, which can rapidly classify the taxonomy of viral sequences at the family level and supports the visualization of the associations of all families. We evaluate the performance of PhaGCN2 and compare it with the state-of-the-art virus classification tools, such as vConTACT2, CAT and VPF-Class, using the widely accepted metrics. The results show that PhaGCN2 largely improves the precision and recall of virus classification, increases the number of classifiable virus sequences in the Global Ocean Virome dataset (v2.0) by four times and classifies more than 90% of the Gut Phage Database. PhaGCN2 makes it possible to conduct high-throughput and automatic expansion of the database of the International Committee on Taxonomy of Viruses. The source code is freely available at https://github.com/KennthShang/PhaGCN2.0.

The arrival of the Infinium DNA methylation BeadChips for mice and other nonhuman mammalian species has outpaced the development of the informatics that supports their use for epigenetics study in model organisms. Here, we present informatics infrastructure and methods to allow easy DNA methylation analysis on multiple species, including domesticated animals and inbred laboratory mice (in SeSAMe version 1.16.0+). First, we developed a data-driven analysis pipeline covering species inference, genome-specific data preprocessing and regression modeling. We targeted genomes of 310 species and 37 inbred mouse strains and showed that genome-specific preprocessing prevents artifacts and yields more accurate measurements than generic pipelines. Second, we uncovered the dynamics of the epigenome evolution in different genomic territories and tissue types through comparative analysis. We identified a catalog of inbred mouse strain-specific methylation differences, some of which are linked to the strains' immune, metabolic and neurological phenotypes. By streamlining DNA methylation array analysis for undesigned genomes, our methods extend epigenome research to broad species contexts.

Accumulating evidences demonstrate that circular RNA (circRNA) plays an important role in human diseases. Identification of circRNA-disease associations can help for the diagnosis of human diseases, while the traditional method based on biological experiments is time-consuming. In order to address the limitation, a series of computational methods have been proposed in recent years. However, few works have summarized these methods or compared the performance of them. In this paper, we divided the existing methods into three categories: information propagation, traditional machine learning and deep learning. Then, the baseline methods in each category are introduced in detail. Further, 5 different datasets are collected, and 14 representative methods of each category are selected and compared in the 5-fold, 10-fold cross-validation and the de novo experiment. In order to further evaluate the effectiveness of these methods, six common cancers are selected to compare the number of correctly identified circRNA-disease associations in the top-10, top-20, top-50, top-100 and top-200. In addition, according to the results, the observation about the robustness and the character of these methods are concluded. Finally, the future directions and challenges are discussed.

Precisely predicting the drug-drug interaction (DDI) is an important application and host research topic in drug discovery, especially for avoiding the adverse effect when using drug combination treatment for patients. Nowadays, machine learning and deep learning methods have achieved great success in DDI prediction. However, we notice that most of the works ignore the importance of the relation type when building the DDI prediction models. In this work, we propose a novel R$^2$-DDI framework, which introduces a relation-aware feature refinement module for drug representation learning. The relation feature is integrated into drug representation and refined in the framework. With the refinement features, we also incorporate the consistency training method to regularize the multi-branch predictions for better generalization. Through extensive experiments and studies, we demonstrate our R$^2$-DDI approach can significantly improve the DDI prediction performance over multiple real-world datasets and settings, and our method shows better generalization ability with the help of the feature refinement design.

In the current era, one of the major challenges is to manage the treatment of drug/antibiotic-resistant strains of bacteria. Phage therapy, a century-old technique, may serve as an alternative to antibiotics in treating bacterial infections caused by drug-resistant strains of bacteria. In this review, a systematic attempt has been made to summarize phage-based therapy in depth. This review has been divided into the following two sections: general information and computer-aided phage therapy (CAPT). In the case of general information, we cover the history of phage therapy, the mechanism of action, the status of phage-based products (approved and clinical trials) and the challenges. This review emphasizes CAPT, where we have covered primary phage-associated resources, phage prediction methods and pipelines. This review covers a wide range of databases and resources, including viral genomes and proteins, phage receptors, host genomes of phages, phage-host interactions and lytic proteins. In the post-genomic era, identifying the most suitable phage for lysing a drug-resistant strain of bacterium is crucial for developing alternate treatments for drug-resistant bacteria and this remains a challenging problem. Thus, we compile all phage-associated prediction methods that include the prediction of phages for a bacterial strain, the host for a phage and the identification of interacting phage-host pairs. Most of these methods have been developed using machine learning and deep learning techniques. This review also discussed recent advances in the field of CAPT, where we briefly describe computational tools available for predicting phage virions, the life cycle of phages and prophage identification. Finally, we describe phage-based therapy's advantages, challenges and opportunities.

In recent years, knowledge graphs (KGs) have gained a great deal of popularity as a tool for storing relationships between entities and for performing higher level reasoning. KGs in biomedicine and clinical practice aim to provide an elegant solution for diagnosing and treating complex diseases more efficiently and flexibly. Here, we provide a systematic review to characterize the state-of-the-art of KGs in the area of complex disease research. We cover the following topics: (1) knowledge sources, (2) entity extraction methods, (3) relation extraction methods and (4) the application of KGs in complex diseases. As a result, we offer a complete picture of the domain. Finally, we discuss the challenges in the field by identifying gaps and opportunities for further research and propose potential research directions of KGs for complex disease diagnosis and treatment.

Bacteriophages (or phages), which infect bacteria, have two distinct lifestyles: virulent and temperate. Predicting the lifestyle of phages helps decipher their interactions with their bacterial hosts, aiding phages' applications in fields such as phage therapy. Because experimental methods for annotating the lifestyle of phages cannot keep pace with the fast accumulation of sequenced phages, computational method for predicting phages' lifestyles has become an attractive alternative. Despite some promising results, computational lifestyle prediction remains difficult because of the limited known annotations and the sheer amount of sequenced phage contigs assembled from metagenomic data. In particular, most of the existing tools cannot precisely predict phages' lifestyles for short contigs. In this work, we develop PhaTYP (Phage TYPe prediction tool) to improve the accuracy of lifestyle prediction on short contigs. We design two different training tasks, self-supervised and fine-tuning tasks, to overcome lifestyle prediction difficulties. We rigorously tested and compared PhaTYP with four state-of-the-art methods: DeePhage, PHACTS, PhagePred and BACPHLIP. The experimental results show that PhaTYP outperforms all these methods and achieves more stable performance on short contigs. In addition, we demonstrated the utility of PhaTYP for analyzing the phage lifestyle on human neonates' gut data. This application shows that PhaTYP is a useful means for studying phages in metagenomic data and helps extend our understanding of microbial communities.

Drug-target interaction (DTI) prediction is an essential step in drug repositioning. A few graph neural network (GNN)-based methods have been proposed for DTI prediction using heterogeneous biological data. However, existing GNN-based methods only aggregate information from directly connected nodes restricted in a drug-related or a target-related network and are incapable of capturing high-order dependencies in the biological heterogeneous graph. In this paper, we propose a metapath-aggregated heterogeneous graph neural network (MHGNN) to capture complex structures and rich semantics in the biological heterogeneous graph for DTI prediction. Specifically, MHGNN enhances heterogeneous graph structure learning and high-order semantics learning by modeling high-order relations via metapaths. Additionally, MHGNN enriches high-order correlations between drug-target pairs (DTPs) by constructing a DTP correlation graph with DTPs as nodes. We conduct extensive experiments on three biological heterogeneous datasets. MHGNN favorably surpasses 17 state-of-the-art methods over 6 evaluation metrics, which verifies its efficacy for DTI prediction. The code is available at https://github.com/Zora-LM/MHGNN-DTI.

Drug-target binding affinity prediction is a fundamental task for drug discovery and has been studied for decades. Most methods follow the canonical paradigm that processes the inputs of the protein (target) and the ligand (drug) separately and then combines them together. In this study we demonstrate, surprisingly, that a model is able to achieve even superior performance without access to any protein-sequence-related information. Instead, a protein is characterized completely by the ligands that it interacts. Specifically, we treat different proteins separately, which are jointly trained in a multi-head manner, so as to learn a robust and universal representation of ligands that is generalizable across proteins. Empirical evidences show that the novel paradigm outperforms its competitive sequence-based counterpart, with the Mean Squared Error (MSE) of 0.4261 versus 0.7612 and the R-Square of 0.7984 versus 0.6570 compared with DeepAffinity. We also investigate the transfer learning scenario where unseen proteins are encountered after the initial training, and the cross-dataset evaluation for prospective studies. The results reveals the robustness of the proposed model in generalizing to unseen proteins as well as in predicting future data. Source codes and data are available at https://github.com/huzqatpku/SAM-DTA.

Single-cell analysis is a valuable approach for dissecting the cellular heterogeneity, and single-cell chromatin accessibility sequencing (scCAS) can profile the epigenetic landscapes for thousands of individual cells. It is challenging to analyze scCAS data, because of its high dimensionality and a higher degree of sparsity compared with scRNA-seq data. Topic modeling in single-cell data analysis can lead to robust identification of the cell types and it can provide insight into the regulatory mechanisms. Reference-guided approach may facilitate the analysis of scCAS data by utilizing the information in existing datasets. We present RefTM (Reference-guided Topic Modeling of single-cell chromatin accessibility data), which not only utilizes the information in existing bulk chromatin accessibility and annotated scCAS data, but also takes advantage of topic models for single-cell data analysis. RefTM simultaneously models: (1) the shared biological variation among reference data and the target scCAS data; (2) the unique biological variation in scCAS data; (3) other variations from known covariates in scCAS data.

Spatially resolved transcriptomics technologies enable the measurement of transcriptome information while retaining the spatial context at the regional, cellular or sub-cellular level. While previous computational methods have relied on gene expression information alone for clustering single-cell populations, more recent methods have begun to leverage spatial location and histology information to improve cell clustering and cell-type identification. In this study, using seven semi-synthetic datasets with real spatial locations, simulated gene expression and histology images as well as ground truth cell-type labels, we evaluate 15 clustering methods based on clustering accuracy, robustness to data variation and input parameters, computational efficiency, and software usability. Our analysis demonstrates that even though incorporating the additional spatial and histology information leads to increased accuracy in some datasets, it does not consistently improve clustering compared with using only gene expression data. Our results indicate that for the clustering of spatial transcriptomics data, there are still opportunities to enhance the overall accuracy and robustness by improving information extraction and feature selection from spatial and histology data.

MOTIVATION: Single-cell assay for transposase accessible chromatin using sequencing (scATAC-seq) is a valuable resource to learn cis-regulatory elements such as cell-type specific enhancers and transcription factor binding sites. However, cell-type identification of scATAC-seq data is known to be challenging due to the heterogeneity derived from different protocols and the high dropout rate. RESULTS: In this study, we perform a systematic comparison of seven scATAC-seq datasets of mouse brain to benchmark the efficacy of neuronal cell-type annotation from gene sets. We find that redundant marker genes give a dramatic improvement for a sparse scATAC-seq annotation across the data collected from different studies. Interestingly, simple aggregation of such marker genes achieves performance comparable or higher than that of machine-learning classifiers, suggesting its potential for downstream applications. Based on our results, we reannotated all scATAC-seq data for detailed cell types using robust marker genes. Their meta scATAC-seq profiles are publicly available at https://gillisweb.cshl.edu/Meta_scATAC. Furthermore, we trained a deep neural network to predict chromatin accessibility from only DNA sequence and identified key motifs enriched for each neuronal subtype. Those predicted profiles are visualized together in our database as a valuable resource to explore cell-type specific epigenetic regulation in a sequence-dependent and -independent manner.

Deciphering 3D genome conformation is important for understanding gene regulation and cellular function at a spatial level. The recent advances of single cell Hi-C technologies have enabled the profiling of the 3D architecture of DNA within individual cell, which allows us to study the cell-to-cell variability of 3D chromatin organization. Computational approaches are in urgent need to comprehensively analyze the sparse and heterogeneous single cell Hi-C data. Here, we proposed scDEC-Hi-C, a new framework for single cell Hi-C analysis with deep generative neural networks. scDEC-Hi-C outperforms existing methods in terms of single cell Hi-C data clustering and imputation. Moreover, the generative power of scDEC-Hi-C could help unveil the differences of chromatin architecture across cell types. We expect that scDEC-Hi-C could shed light on deepening our understanding of the complex mechanism underlying the formation of chromatin contacts.

Differential abundance analysis (DAA) is one central statistical task in microbiome data analysis. A robust and powerful DAA tool can help identify highly confident microbial candidates for further biological validation. Current microbiome studies frequently generate correlated samples from different microbiome sampling schemes such as spatial and temporal sampling. In the past decade, a number of DAA tools for correlated microbiome data (DAA-c) have been proposed. Disturbingly, different DAA-c tools could sometimes produce quite discordant results. To recommend the best practice to the field, we performed the first comprehensive evaluation of existing DAA-c tools using real data-based simulations. Overall, the linear model-based methods LinDA, MaAsLin2 and LDM are more robust than methods based on generalized linear models. The LinDA method is the only method that maintains reasonable performance in the presence of strong compositional effects.

Transcriptome signature reversion (TSR) has been extensively proposed and used to discover new indications for existing drugs (i.e. drug repositioning, drug repurposing) for various cancer types. TSR relies on the assumption that a drug that can revert gene expression changes induced by a disease back to original, i.e. healthy, levels is likely to be therapeutically active in treating the disease. Here, we aimed to validate the concept of TSR using the PRISM repurposing data set, which is-as of writing-the largest pharmacogenomic data set. The predictive utility of the TSR approach as it has currently been used appears to be much lower than previously reported and is completely nullified after the drug gene expression signatures are adjusted for the general anti-proliferative downstream effects of drug-induced decreased cell viability. Therefore, TSR mainly relies on generic anti-proliferative drug effects rather than on targeting cancer pathways specifically upregulated in tumor types.

Genome-wide association studies (GWAS) have benefited greatly from enhanced high-throughput technology in recent decades. GWAS meta-analysis has become increasingly popular to highlight the genetic architecture of complex traits, informing about the replicability and variability of effect estimations across human ancestries. A wealth of GWAS meta-analysis methodologies have been developed depending on the input data and the outcome information of interest. We present a survey of current approaches from SNP to pathway-based meta-analysis by acknowledging the range of resources and methodologies in the field, and we provide a comprehensive review of different categories of Genome-Wide Meta-analysis methods employed. These methods highlight different levels at which GWAS meta-analysis may be done, including Single Nucleotide Polymorphisms, Genes and Pathways, for which we describe their framework outline. We also discuss the strengths and pitfalls of each approach and make suggestions regarding each of them.

DNA and RNA sequencing technologies have revolutionized biology and biomedical sciences, sequencing full genomes and transcriptomes at very high speeds and reasonably low costs. RNA sequencing (RNA-Seq) enables transcript identification and quantification, but once sequencing has concluded researchers can be easily overwhelmed with questions such as how to go from raw data to differential expression (DE), pathway analysis and interpretation. Several pipelines and procedures have been developed to this effect. Even though there is no unique way to perform RNA-Seq analysis, it usually follows these steps: 1) raw reads quality check, 2) alignment of reads to a reference genome, 3) aligned reads' summarization according to an annotation file, 4) DE analysis and 5) gene set analysis and/or functional enrichment analysis. Each step requires researchers to make decisions, and the wide variety of options and resulting large volumes of data often lead to interpretation challenges. There also seems to be insufficient guidance on how best to obtain relevant information and derive actionable knowledge from transcription experiments. In this paper, we explain RNA-Seq steps in detail and outline differences and similarities of different popular options, as well as advantages and disadvantages. We also discuss non-coding RNA analysis, multi-omics, meta-transcriptomics and the use of artificial intelligence methods complementing the arsenal of tools available to researchers. Lastly, we perform a complete analysis from raw reads to DE and functional enrichment analysis, visually illustrating how results are not absolute truths and how algorithmic decisions can greatly impact results and interpretation.

Single-cell RNA sequencing (scRNA-seq) clustering and labelling methods are used to determine precise cellular composition of tissue samples. Automated labelling methods rely on either unsupervised, cluster-based approaches or supervised, cell-based approaches to identify cell types. The high complexity of cancer poses a unique challenge, as tumor microenvironments are often composed of diverse cell subpopulations with unique functional effects that may lead to disease progression, metastasis and treatment resistance. Here, we assess 17 cell-based and 9 cluster-based scRNA-seq labelling algorithms using 8 cancer datasets, providing a comprehensive large-scale assessment of such methods in a cancer-specific context. Using several performance metrics, we show that cell-based methods generally achieved higher performance and were faster compared to cluster-based methods. Cluster-based methods more successfully labelled non-malignant cell types, likely because of a lack of gene signatures for relevant malignant cell subpopulations. Larger cell numbers present in some cell types in training data positively impacted prediction scores for cell-based methods. Finally, we examined which methods performed favorably when trained and tested on separate patient cohorts in scenarios similar to clinical applications, and which were able to accurately label particularly small or under-represented cell populations in the given datasets. We conclude that scPred and SVM show the best overall performances with cancer-specific data and provide further suggestions for algorithm selection. Our analysis pipeline for assessing the performance of cell type labelling algorithms is available in https://github.com/shooshtarilab/scRNAseq-Automated-Cell-Type-Labelling.

Although previous studies have revealed that synonymous mutations contribute to various human diseases, distinguishing deleterious synonymous mutations from benign ones is still a challenge in medical genomics. Recently, computational tools have been introduced to predict the harmfulness of synonymous mutations. However, most of these computational tools rely on balanced training sets without considering abundant negative samples that could result in deficient performance. In this study, we propose a computational model that uses a selective ensemble to predict deleterious synonymous mutations (seDSM). We construct several candidate base classifiers for the ensemble using balanced training subsets randomly sampled from the imbalanced benchmark training sets. The diversity measures of the base classifiers are calculated by the pairwise diversity metrics, and the classifiers with the highest diversities are selected for integration using soft voting for synonymous mutation prediction. We also design two strategies for filling in missing values in the imbalanced dataset and constructing models using different pairwise diversity metrics. The experimental results show that a selective ensemble based on double fault with the ensemble strategy EKNNI for filling in missing values is the most effective scheme. Finally, using 40-dimensional biology features, we propose a novel model based on a selective ensemble for predicting deleterious synonymous mutations (seDSM). seDSM outperformed other state-of-the-art methods on the independent test sets according to multiple evaluation indicators, indicating that it has an outstanding predictive performance for deleterious synonymous mutations. We hope that seDSM will be useful for studying deleterious synonymous mutations and advancing our understanding of synonymous mutations. The source code of seDSM is freely accessible at https://github.com/xialab-ahu/seDSM.git.

Cell-cell communications are vital for biological signalling and play important roles in complex diseases. Recent advances in single-cell spatial transcriptomics (SCST) technologies allow examining the spatial cell communication landscapes and hold the promise for disentangling the complex ligand-receptor (L-R) interactions across cells. However, due to frequent dropout events and noisy signals in SCST data, it is challenging and lack of effective and tailored methods to accurately infer cellular communications. Herein, to decipher the cell-to-cell communications from SCST profiles, we propose a novel adaptive graph model with attention mechanisms named spaCI. spaCI incorporates both spatial locations and gene expression profiles of cells to identify the active L-R signalling axis across neighbouring cells. Through benchmarking with currently available methods, spaCI shows superior performance on both simulation data and real SCST datasets. Furthermore, spaCI is able to identify the upstream transcriptional factors mediating the active L-R interactions. For biological insights, we have applied spaCI to the seqFISH+ data of mouse cortex and the NanoString CosMx Spatial Molecular Imager (SMI) data of non-small cell lung cancer samples. spaCI reveals the hidden L-R interactions from the sparse seqFISH+ data, meanwhile identifies the inconspicuous L-R interactions including THBS1-ITGB1 between fibroblast and tumours in NanoString CosMx SMI data. spaCI further reveals that SMAD3 plays an important role in regulating the crosstalk between fibroblasts and tumours, which contributes to the prognosis of lung cancer patients. Collectively, spaCI addresses the challenges in interrogating SCST data for gaining insights into the underlying cellular communications, thus facilitates the discoveries of disease mechanisms, effective biomarkers and therapeutic targets.

With the development of genome sequencing technology, using computing technology to predict grain protein function has become one of the important tasks of bioinformatics. The protein data of four grains, soybean, maize, indica and japonica are selected in this experimental dataset. In this paper, a novel neural network algorithm Chemical-SA-BiLSTM is proposed for grain protein function prediction. The Chemical-SA-BiLSTM algorithm fuses the chemical properties of proteins on the basis of amino acid sequences, and combines the self-attention mechanism with the bidirectional Long Short-Term Memory network. The experimental results show that the Chemical-SA-BiLSTM algorithm is superior to other classical neural network algorithms, and can more accurately predict the protein function, which proves the effectiveness of the Chemical-SA-BiLSTM algorithm in the prediction of grain protein function. The source code of our method is available at https://github.com/HwaTong/Chemical-SA-BiLSTM.

Aberrant DNA methylation is the most common molecular lesion that is crucial for the occurrence and development of cancer, but has thus far been underappreciated as a clinical tool for cancer classification, diagnosis or as a guide for therapeutic decisions. Partly, this has been due to a lack of proven algorithms that can use methylation data to stratify patients into clinically relevant risk groups and subtypes that are of prognostic importance. Here, we proposed a novel Bayesian model to capture the methylation signatures of different subtypes from paired normal and tumor methylation array data. Application of our model to synthetic and empirical data showed high clustering accuracy, and was able to identify the possible epigenetic cause of a cancer subtype.

Medicinal plants are the main source of natural metabolites with specialised pharmacological activities and have been widely examined by plant researchers. Numerous omics studies of medicinal plants have been performed to identify molecular markers of species and functional genes controlling key biological traits, as well as to understand biosynthetic pathways of bioactive metabolites and the regulatory mechanisms of environmental responses. Omics technologies have been widely applied to medicinal plants, including as taxonomics, transcriptomics, metabolomics, proteomics, genomics, pangenomics, epigenomics and mutagenomics. However, because of the complex biological regulation network, single omics usually fail to explain the specific biological phenomena. In recent years, reports of integrated multi-omics studies of medicinal plants have increased. Until now, there have few assessments of recent developments and upcoming trends in omics studies of medicinal plants. We highlight recent developments in omics research of medicinal plants, summarise the typical bioinformatics resources available for analysing omics datasets, and discuss related future directions and challenges. This information facilitates further studies of medicinal plants, refinement of current approaches and leads to new ideas.

Identifying cancer type-specific driver mutations is crucial for illuminating distinct pathologic mechanisms across various tumors and providing opportunities of patient-specific treatment. However, although many computational methods were developed to predict driver mutations in a type-specific manner, the methods still have room to improve. Here, we devise a novel feature based on sequence co-evolution analysis to identify cancer type-specific driver mutations and construct a machine learning (ML) model with state-of-the-art performance. Specifically, relying on 28 000 tumor samples across 66 cancer types, our ML framework outperformed current leading methods of detecting cancer driver mutations. Interestingly, the cancer mutations identified by sequence co-evolution feature are frequently observed in interfaces mediating tissue-specific protein-protein interactions that are known to associate with shaping tissue-specific oncogenesis. Moreover, we provide pre-calculated potential oncogenicity on available human proteins with prediction scores of all possible residue alterations through user-friendly website (http://sbi.postech.ac.kr/w/cancerCE). This work will facilitate the identification of cancer type-specific driver mutations in newly sequenced tumor samples.

The proliferation of single-cell multimodal sequencing technologies has enabled us to understand cellular heterogeneity with multiple views, providing novel and actionable biological insights into the disease-driving mechanisms. Here, we propose a comprehensive end-to-end single-cell multimodal analysis framework named Deep Parametric Inference (DPI). DPI transforms single-cell multimodal data into a multimodal parameter space by inferring individual modal parameters. Analysis of cord blood mononuclear cells (CBMC) reveals that the multimodal parameter space can characterize the heterogeneity of cells more comprehensively than individual modalities. Furthermore, comparisons with the state-of-the-art methods on multiple datasets show that DPI has superior performance. Additionally, DPI can reference and query cell types without batch effects. As a result, DPI can successfully analyze the progression of COVID-19 disease in peripheral blood mononuclear cells (PBMC). Notably, we further propose a cell state vector field and analyze the transformation pattern of bone marrow cells (BMC) states. In conclusion, DPI is a powerful single-cell multimodal analysis framework that can provide new biological insights into biomedical researchers. The python packages, datasets and user-friendly manuals of DPI are freely available at https://github.com/studentiz/dpi.

Recently, peptide-based drugs have gained unprecedented interest in discovering and developing antifungal drugs due to their high efficacy, broad-spectrum activity, low toxicity and few side effects. However, it is time-consuming and expensive to identify antifungal peptides (AFPs) experimentally. Therefore, computational methods for accurately predicting AFPs are highly required. In this work, we develop AFP-MFL, a novel deep learning model that predicts AFPs only relying on peptide sequences without using any structural information. AFP-MFL first constructs comprehensive feature profiles of AFPs, including contextual semantic information derived from a pre-trained protein language model, evolutionary information, and physicochemical properties. Subsequently, the co-attention mechanism is utilized to integrate contextual semantic information with evolutionary information and physicochemical properties separately. Extensive experiments show that AFP-MFL outperforms state-of-the-art models on four independent test datasets. Furthermore, the SHAP method is employed to explore each feature contribution to the AFPs prediction. Finally, a user-friendly web server of the proposed AFP-MFL is developed and freely accessible at http://inner.wei-group.net/AFPMFL/, which can be considered as a powerful tool for the rapid screening and identification of novel AFPs.

Discovering the relationships between long non-coding RNAs (lncRNAs) and diseases is significant in the treatment, diagnosis and prevention of diseases. However, current identified lncRNA-disease associations are not enough because of the expensive and heavy workload of wet laboratory experiments. Therefore, it is greatly important to develop an efficient computational method for predicting potential lncRNA-disease associations. Previous methods showed that combining the prediction results of the lncRNA-disease associations predicted by different classification methods via Learning to Rank (LTR) algorithm can be effective for predicting potential lncRNA-disease associations. However, when the classification results are incorrect, the ranking results will inevitably be affected. We propose the GraLTR-LDA predictor based on biological knowledge graphs and ranking framework for predicting potential lncRNA-disease associations. Firstly, homogeneous graph and heterogeneous graph are constructed by integrating multi-source biological information. Then, GraLTR-LDA integrates graph auto-encoder and attention mechanism to extract embedded features from the constructed graphs. Finally, GraLTR-LDA incorporates the embedded features into the LTR via feature crossing statistical strategies to predict priority order of diseases associated with query lncRNAs. Experimental results demonstrate that GraLTR-LDA outperforms the other state-of-the-art predictors and can effectively detect potential lncRNA-disease associations. Availability and implementation: Datasets and source codes are available at http://bliulab.net/GraLTR-LDA.

This paper describes a method Pprint2, which is an improved version of Pprint developed for predicting RNA-interacting residues in a protein. Training and independent/validation datasets used in this study comprises of 545 and 161 non-redundant RNA-binding proteins, respectively. All models were trained on training dataset and evaluated on the validation dataset. The preliminary analysis reveals that positively charged amino acids such as H, R and K, are more prominent in the RNA-interacting residues. Initially, machine learning based models have been developed using binary profile and obtain maximum area under curve (AUC) 0.68 on validation dataset. The performance of this model improved significantly from AUC 0.68 to 0.76, when evolutionary profile is used instead of binary profile. The performance of our evolutionary profile-based model improved further from AUC 0.76 to 0.82, when convolutional neural network has been used for developing model. Our final model based on convolutional neural network using evolutionary information achieved AUC 0.82 with Matthews correlation coefficient of 0.49 on the validation dataset. Our best model outperforms existing methods when evaluated on the independent/validation dataset. A user-friendly standalone software and web-based server named 'Pprint2' has been developed for predicting RNA-interacting residues (https://webs.iiitd.edu.in/raghava/pprint2 and https://github.com/raghavagps/pprint2).

Bulk sequencing experiments (single- and multi-omics) are essential for exploring wide-ranging biological questions. To facilitate interactive, exploratory tasks, coupled with the sharing of easily accessible information, we present bulkAnalyseR, a package integrating state-of-the-art approaches using an expression matrix as the starting point (pre-processing functions are available as part of the package). Static summary images are replaced with interactive panels illustrating quality-checking, differential expression analysis (with noise detection) and biological interpretation (enrichment analyses, identification of expression patterns, followed by inference and comparison of regulatory interactions). bulkAnalyseR can handle different modalities, facilitating robust integration and comparison of cis-, trans- and customised regulatory networks.

Different RNAs have distinct subcellular localizations. However, nucleotide features that determine these distinct distributions of lncRNAs and mRNAs have yet to be fully addressed. Here, we develop RNAlight, a machine learning model based on LightGBM, to identify nucleotide k-mers contributing to the subcellular localizations of mRNAs and lncRNAs. With the Tree SHAP algorithm, RNAlight extracts nucleotide features for cytoplasmic or nuclear localization of RNAs, indicating the sequence basis for distinct RNA subcellular localizations. By assembling k-mers to sequence features and subsequently mapping to known RBP-associated motifs, different types of sequence features and their associated RBPs were additionally uncovered for lncRNAs and mRNAs with distinct subcellular localizations. Finally, we extended RNAlight to precisely predict the subcellular localizations of other types of RNAs, including snRNAs, snoRNAs and different circular RNA transcripts, suggesting the generality of using RNAlight for RNA subcellular localization prediction.

Accurately predicting the interaction modes for metalloproteins remains extremely challenging in structure-based drug design and mechanism analysis of enzymatic catalysis due to the complexity of metal coordination in metalloproteins. Here, we report a docking method for metalloproteins based on geometric probability (GPDOCK) with unprecedented accuracy. The docking tests of 10 common metal ions with 9360 metalloprotein-ligand complexes demonstrate that GPDOCK has an accuracy of 94.3% in predicting binding pose. What is more, it can accurately realize the docking of metalloproteins with ligand when one or two water molecules are engaged in the metal ion coordination. Since GPDOCK only depends on the three-dimensional structure of metalloprotein and ligand, structure-based machine learning model is employed for the scoring of binding poses, which significantly improves computational efficiency. The proposed docking strategy can be an effective and efficient tool for drug design and further study of binding mechanism of metalloproteins. The manual of GPDOCK and the code for the logistical regression model used to re-rank the docking results are available at https://github.com/wangkai-zhku/GPDOCK.git.

Drug response prediction in cancer cell lines is of great significance in personalized medicine. In this study, we propose GADRP, a cancer drug response prediction model based on graph convolutional networks (GCNs) and autoencoders (AEs). We first use a stacked deep AE to extract low-dimensional representations from cell line features, and then construct a sparse drug cell line pair (DCP) network incorporating drug, cell line, and DCP similarity information. Later, initial residual and layer attention-based GCN (ILGCN) that can alleviate over-smoothing problem is utilized to learn DCP features. And finally, fully connected network is employed to make prediction. Benchmarking results demonstrate that GADRP can significantly improve prediction performance on all metrics compared with baselines on five datasets. Particularly, experiments of predictions of unknown DCP responses, drug-cancer tissue associations， and drug-pathway associations illustrate the predictive power of GADRP. All results highlight the effectiveness of GADRP in predicting drug responses, and its potential value in guiding anti-cancer drug selection.

Qualitative or quantitative prediction models of structure-activity relationships based on graph neural networks (GNNs) are prevalent in drug discovery applications and commonly have excellently predictive power. However, the network information flows of GNNs are highly complex and accompanied by poor interpretability. Unfortunately, there are relatively less studies on GNN attributions, and their developments in drug research are still at the early stages. In this work, we adopted several advanced attribution techniques for different GNN frameworks and applied them to explain multiple drug molecule property prediction tasks, enabling the identification and visualization of vital chemical information in the networks. Additionally, we evaluated them quantitatively with attribution metrics such as accuracy, sparsity, fidelity and infidelity, stability and sensitivity; discussed their applicability and limitations; and provided an open-source benchmark platform for researchers. The results showed that all attribution techniques were effective, while those directly related to the predicted labels, such as integrated gradient, preferred to have better attribution performance. These attribution techniques we have implemented could be directly used for the vast majority of chemical GNN interpretation tasks.

Understanding the mechanisms of candidate drugs play an important role in drug discovery. The activating/inhibiting mechanisms between drugs and targets are major types of mechanisms of drugs. Owing to the complexity of drug-target (DT) mechanisms and data scarcity, modelling this problem based on deep learning methods to accurately predict DT activating/inhibiting mechanisms remains a considerable challenge. Here, by considering network pharmacology, we propose a multi-view deep learning model, DrugAI, which combines four modules, i.e. a graph neural network for drugs, a convolutional neural network for targets, a network embedding module for drugs and targets and a deep neural network for predicting activating/inhibiting mechanisms between drugs and targets. Computational experiments show that DrugAI performs better than state-of-the-art methods and has good robustness and generalization. To demonstrate the reliability of the predictive results of DrugAI, bioassay experiments are conducted to validate two drugs (notopterol and alpha-asarone) predicted to activate TRPV1. Moreover, external validation bears out 61 pairs of mechanism relationships between natural products and their targets predicted by DrugAI based on independent literatures and PubChem bioassays. DrugAI, for the first time, provides a powerful multi-view deep learning framework for robust prediction of DT activating/inhibiting mechanisms.

Protein-protein interactions play an important role in many biological processes. However, although structure prediction for monomer proteins has achieved great progress with the advent of advanced deep learning algorithms like AlphaFold, the structure prediction for protein-protein complexes remains an open question. Taking advantage of the Transformer model of ESM-MSA, we have developed a deep learning-based model, named DeepHomo2.0, to predict protein-protein interactions of homodimeric complexes by leveraging the direct-coupling analysis (DCA) and Transformer features of sequences and the structure features of monomers. DeepHomo2.0 was extensively evaluated on diverse test sets and compared with eight state-of-the-art methods including protein language model-based, DCA-based and machine learning-based methods. It was shown that DeepHomo2.0 achieved a high precision of >70% with experimental monomer structures and >60% with predicted monomer structures for the top 10 predicted contacts on the test sets and outperformed the other eight methods. Moreover, even the version without using structure information, named DeepHomoSeq, still achieved a good precision of >55% for the top 10 predicted contacts. Integrating the predicted contacts into protein docking significantly improved the structure prediction of realistic Critical Assessment of Protein Structure Prediction homodimeric complexes. DeepHomo2.0 and DeepHomoSeq are available at http://huanglab.phys.hust.edu.cn/DeepHomo2/.

As one of the most vital methods in drug development, drug repositioning emphasizes further analysis and research of approved drugs based on the existing large amount of clinical and experimental data to identify new indications of drugs. However, the existing drug repositioning methods didn't achieve enough prediction performance, and these methods do not consider the effectiveness information of drugs, which make it difficult to obtain reliable and valuable results. In this study, we proposed a drug repositioning framework termed DRONet, which make full use of effectiveness comparative relationships (ECR) among drugs as prior information by combining network embedding and ranking learning. We utilized network embedding methods to learn the deep features of drugs from a heterogeneous drug-disease network, and constructed a high-quality drug-indication data set including effectiveness-based drug contrast relationships. The embedding features and ECR of drugs are combined effectively through a designed ranking learning model to prioritize candidate drugs. Comprehensive experiments show that DRONet has higher prediction accuracy (improving 87.4% on Hit@1 and 37.9% on mean reciprocal rank) than state of the art. The case analysis also demonstrates high reliability of predicted results, which has potential to guide clinical drug development.

Membrane-based cells are the fundamental structural and functional units of organisms, while evidences demonstrate that liquid-liquid phase separation (LLPS) is associated with the formation of membraneless organelles, such as P-bodies, nucleoli and stress granules. Many studies have been undertaken to explore the functions of protein phase separation (PS), but these studies lacked an effective tool to identify the sequence segments that critical for LLPS. In this study, we presented a novel software called dSCOPE (http://dscope.omicsbio.info) to predict the PS-driving regions. To develop the predictor, we curated experimentally identified sequence segments that can drive LLPS from published literature. Then sliding sequence window based physiological, biochemical, structural and coding features were integrated by random forest algorithm to perform prediction. Through rigorous evaluation, dSCOPE was demonstrated to achieve satisfactory performance. Furthermore, large-scale analysis of human proteome based on dSCOPE showed that the predicted PS-driving regions enriched various protein post-translational modifications and cancer mutations, and the proteins which contain predicted PS-driving regions enriched critical cellular signaling pathways. Taken together, dSCOPE precisely predicted the protein sequence segments critical for LLPS, with various helpful information visualized in the webserver to facilitate LLPS-related research.

Numerous experimental studies have indicated that alteration and dysregulation in mircroRNAs (miRNAs) are associated with serious diseases. Identifying disease-related miRNAs is therefore an essential and challenging task in bioinformatics research. Computational methods are an efficient and economical alternative to conventional biomedical studies and can reveal underlying miRNA-disease associations for subsequent experimental confirmation with reasonable confidence. Despite the success of existing computational approaches, most of them only rely on the known miRNA-disease associations to predict associations without adding other data to increase the prediction accuracy, and they are affected by issues of data sparsity. In this paper, we present MRRN, a model that combines matrix reconstruction with node reliability to predict probable miRNA-disease associations. In MRRN, the most reliable neighbors of miRNA and disease are used to update the original miRNA-disease association matrix, which significantly reduces data sparsity. Unknown miRNA-disease associations are reconstructed by aggregating the most reliable first-order neighbors to increase prediction accuracy by representing the local and global structure of the heterogeneous network. Five-fold cross-validation of MRRN produced an area under the curve (AUC) of 0.9355 and area under the precision-recall curve (AUPR) of 0.2646, values that were greater than those produced by comparable models. Two different types of case studies using three diseases were conducted to demonstrate the accuracy of MRRN, and all top 30 predicted miRNAs were verified.

Long non-coding RNAs (lncRNAs) played essential roles in nearly every biological process and disease. Many algorithms were developed to distinguish lncRNAs from mRNAs in transcriptomic data and facilitated discoveries of more than 600 000 of lncRNAs. However, only a tiny fraction (<1%) of lncRNA transcripts (~4000) were further validated by low-throughput experiments (EVlncRNAs). Given the cost and labor-intensive nature of experimental validations, it is necessary to develop computational tools to prioritize those potentially functional lncRNAs because many lncRNAs from high-throughput sequencing (HTlncRNAs) could be resulted from transcriptional noises. Here, we employed deep learning algorithms to separate EVlncRNAs from HTlncRNAs and mRNAs. For overcoming the challenge of small datasets, we employed a three-layer deep-learning neural network (DNN) with a K-mer feature as the input and a small convolutional neural network (CNN) with one-hot encoding as the input. Three separate models were trained for human (h), mouse (m) and plant (p), respectively. The final concatenated models (EVlncRNA-Dpred (h), EVlncRNA-Dpred (m) and EVlncRNA-Dpred (p)) provided substantial improvement over a previous model based on support-vector-machines (EVlncRNA-pred). For example, EVlncRNA-Dpred (h) achieved 0.896 for the area under receiver-operating characteristic curve, compared with 0.582 given by sequence-based EVlncRNA-pred model. The models developed here should be useful for screening lncRNA transcripts for experimental validations. EVlncRNA-Dpred is available as a web server at https://www.sdklab-biophysics-dzu.net/EVlncRNA-Dpred/index.html, and the data and source code can be freely available along with the web server.

Covalent inhibitors have received extensive attentions in the past few decades because of their long residence time, high binding efficiency and strong selectivity. Therefore, it is valuable to develop computational tools like molecular docking for modeling of covalent protein-ligand interactions or screening of potential covalent drugs. Meeting the needs, we have proposed HCovDock, an efficient docking algorithm for covalent protein-ligand interactions by integrating a ligand sampling method of incremental construction and a scoring function with covalent bond-based energy. Tested on a benchmark containing 207 diverse protein-ligand complexes, HCovDock exhibits a significantly better performance than seven other state-of-the-art covalent docking programs (AutoDock, Cov_DOX, CovDock, FITTED, GOLD, ICM-Pro and MOE). With the criterion of ligand root-mean-squared distance < 2.0 Å, HCovDock obtains a high success rate of 70.5% and 93.2% in reproducing experimentally observed structures for top 1 and top 10 predictions. In addition, HCovDock is also validated in virtual screening against 10 receptors of three proteins. HCovDock is computationally efficient and the average running time for docking a ligand is only 5 min with as fast as 1 sec for ligands with one rotatable bond and about 18 min for ligands with 23 rotational bonds. HCovDock can be freely assessed at http://huanglab.phys.hust.edu.cn/hcovdock/.

Directed protein evolution applies repeated rounds of genetic mutagenesis and phenotypic screening and is often limited by experimental throughput. Through in silico prioritization of mutant sequences, machine learning has been applied to reduce wet lab burden to a level practical for human researchers. On the other hand, robotics permits large batches and rapid iterations for protein engineering cycles, but such capacities have not been well exploited in existing machine learning-assisted directed evolution approaches. Here, we report a scalable and batched method, Bayesian Optimization-guided EVOlutionary (BO-EVO) algorithm, to guide multiple rounds of robotic experiments to explore protein fitness landscapes of combinatorial mutagenesis libraries. We first examined various design specifications based on an empirical landscape of protein G domain B1. Then, BO-EVO was successfully generalized to another empirical landscape of an Escherichia coli kinase PhoQ, as well as simulated NK landscapes with up to moderate epistasis. This approach was then applied to guide robotic library creation and screening to engineer enzyme specificity of RhlA, a key biosynthetic enzyme for rhamnolipid biosurfactants. A 4.8-fold improvement in producing a target rhamnolipid congener was achieved after examining less than 1% of all possible mutants after four iterations. Overall, BO-EVO proves to be an efficient and general approach to guide combinatorial protein engineering without prior knowledge.

MOTIVATION: Exploring the potential long noncoding RNA (lncRNA)-disease associations (LDAs) plays a critical role for understanding disease etiology and pathogenesis. Given the high cost of biological experiments, developing a computational method is a practical necessity to effectively accelerate experimental screening process of candidate LDAs. However, under the high sparsity of LDA dataset, many computational models hardly exploit enough knowledge to learn comprehensive patterns of node representations. Moreover, although the metapath-based GNN has been recently introduced into LDA prediction, it discards intermediate nodes along the meta-path and results in information loss. RESULTS: This paper presents a new multi-view contrastive heterogeneous graph attention network (GAT) for lncRNA-disease association prediction, MCHNLDA for brevity. Specifically, MCHNLDA firstly leverages rich biological data sources of lncRNA, gene and disease to construct two-view graphs, feature structural graph of feature schema view and lncRNA-gene-disease heterogeneous graph of network topology view. Then, we design a cross-contrastive learning task to collaboratively guide graph embeddings of the two views without relying on any labels. In this way, we can pull closer the nodes of similar features and network topology, and push other nodes away. Furthermore, we propose a heterogeneous contextual GAT, where long short-term memory network is incorporated into attention mechanism to effectively capture sequential structure information along the meta-path. Extensive experimental comparisons against several state-of-the-art methods show the effectiveness of proposed framework.The code and data of proposed framework is freely available at https://github.com/zhaoxs686/MCHNLDA.

Differentiating cancer subtypes is crucial to guide personalized treatment and improve the prognosis for patients. Integrating multi-omics data can offer a comprehensive landscape of cancer biological process and provide promising ways for cancer diagnosis and treatment. Taking the heterogeneity of different omics data types into account, we propose a hierarchical multi-kernel learning (hMKL) approach, a novel cancer molecular subtyping method to identify cancer subtypes by adopting a two-stage kernel learning strategy. In stage 1, we obtain a composite kernel borrowing the cancer integration via multi-kernel learning (CIMLR) idea by optimizing the kernel parameters for individual omics data type. In stage 2, we obtain a final fused kernel through a weighted linear combination of individual kernels learned from stage 1 using an unsupervised multiple kernel learning method. Based on the final fusion kernel, k-means clustering is applied to identify cancer subtypes. Simulation studies show that hMKL outperforms the one-stage CIMLR method when there is data heterogeneity. hMKL can estimate the number of clusters correctly, which is the key challenge in subtyping. Application to two real data sets shows that hMKL identified meaningful subtypes and key cancer-associated biomarkers. The proposed method provides a novel toolkit for heterogeneous multi-omics data integration and cancer subtypes identification.

Identification of RNA-small molecule binding sites plays an essential role in RNA-targeted drug discovery and development. These small molecules are expected to be leading compounds to guide the development of new types of RNA-targeted therapeutics compared with regular therapeutics targeting proteins. RNAs can provide many potential drug targets with diverse structures and functions. However, up to now, only a few methods have been proposed. Predicting RNA-small molecule binding sites still remains a big challenge. New computational model is required to better extract the features and predict RNA-small molecule binding sites more accurately. In this paper, a deep learning model, RLBind, was proposed to predict RNA-small molecule binding sites from sequence-dependent and structure-dependent properties by combining global RNA sequence channel and local neighbor nucleotides channel. To our best knowledge, this research was the first to develop a convolutional neural network for RNA-small molecule binding sites prediction. Furthermore, RLBind also can be used as a potential tool when the RNA experimental tertiary structure is not available. The experimental results show that RLBind outperforms other state-of-the-art methods in predicting binding sites. Therefore, our study demonstrates that the combination of global information for full-length sequences and local information for limited local neighbor nucleotides in RNAs can improve the model's predictive performance for binding sites prediction. All datasets and resource codes are available at https://github.com/KailiWang1/RLBind.

Modules consisting of antibiotic resistance genes (ARGs) flanked by inverted repeat Xer-specific recombination sites were thought to be mobile genetic elements that promote horizontal transmission. Less frequently, the presence of mobile modules in plasmids, which facilitate a pdif-mediated ARGs transfer, has been reported. Here, numerous ARGs and toxin-antitoxin genes have been found in pdif site pairs. However, the mechanisms underlying this apparent genetic mobility is currently not understood, and the studies relating to pdif-mediated ARGs transfer onto most bacterial genera are lacking. We developed the web server pdifFinder based on an algorithm called PdifSM that allows the prediction of diverse pdif-ARGs modules in bacterial genomes. Using test set consisting of almost 32 thousand plasmids from 717 species, PdifSM identified 481 plasmids from various bacteria containing pdif sites with ARGs. We found 28-bp-long elements from different genera with clear base preferences. The data we obtained indicate that XerCD-dif site-specific recombination mechanism may have evolutionary adapted to facilitate the pdif-mediated ARGs transfer. Through multiple sequence alignment and evolutionary analyses of duplicated pdif-ARGs modules, we discovered that pdif sites allow an interspecies transfer of ARGs but also across different genera. Mutations in pdif sites generate diverse arrays of modules which mediate multidrug-resistance, as these contain variable numbers of diverse ARGs, insertion sequences and other functional genes. The identification of pdif-ARGs modules and studies focused on the mechanism of ARGs co-transfer will help us to understand and possibly allow controlling the spread of MDR bacteria in clinical settings. The pdifFinder code, standalone software package and description with tutorials are available at https://github.com/mjshao06/pdifFinder.

Analysis of the methylome of tumor cell-free deoxyribonucleic acid (DNA; cfDNA) has emerged as a powerful non-invasive technique for cancer subtyping and prognosis. However, its application is frequently hampered by the quality and total cfDNA yield. Here, we demonstrate the feasibility of very low-input cfDNA for whole-methylome and copy-number profiling studies using enzymatic conversion of unmethylated cysteines [enzymatic methyl-seq (EM-seq)] to better preserve DNA integrity. We created a model for predicting genomic subtyping and prognosis with high accuracy. We validated our tool by comparing whole-genome CpG sequencing with in situ cohorts generated with bisulfite conversion and array hybridization, demonstrating that, despite the different techniques and sample origins, information on cfDNA methylation is comparable with in situ cohorts. Our findings support use of liquid biopsy followed by EM-seq to assess methylome of cancer patients, enabling validation in external cohorts. This advance is particularly relevant for rare cancers like neuroblastomas where liquid-biopsy volume is restricted by ethical regulations in pediatric patients.

Messenger RNA-based therapeutics have shown tremendous potential, as demonstrated by the rapid development of messenger RNA based vaccines for COVID-19. Nevertheless, distribution of mRNA vaccines worldwide has been hampered by mRNA's inherent thermal instability due to in-line hydrolysis, a chemical degradation reaction. Therefore, predicting and understanding RNA degradation is a crucial and urgent task. Here we present RNAdegformer, an effective and interpretable model architecture that excels in predicting RNA degradation. RNAdegformer processes RNA sequences with self-attention and convolutions, two deep learning techniques that have proved dominant in the fields of computer vision and natural language processing, while utilizing biophysical features of RNA. We demonstrate that RNAdegformer outperforms previous best methods at predicting degradation properties at nucleotide resolution for COVID-19 mRNA vaccines. RNAdegformer predictions also exhibit improved correlation with RNA in vitro half-life compared with previous best methods. Additionally, we showcase how direct visualization of self-attention maps assists informed decision-making. Further, our model reveals important features in determining mRNA degradation rates via leave-one-feature-out analysis.

Circular RNAs (circRNAs) are covalently closed transcripts involved in critical regulatory axes, cancer pathways and disease mechanisms. CircRNA expression measured with RNA-seq has particular characteristics that might hamper the performance of standard biostatistical differential expression assessment methods (DEMs). We compared 38 DEM pipelines configured to fit circRNA expression data's statistical properties, including bulk RNA-seq, single-cell RNA-seq (scRNA-seq) and metagenomics DEMs. The DEMs performed poorly on data sets of typical size. Widely used DEMs, such as DESeq2, edgeR and Limma-Voom, gave scarce results, unreliable predictions or even contravened the expected behaviour with some parameter configurations. Limma-Voom achieved the most consistent performance throughout different benchmark data sets and, as well as SAMseq, reasonably balanced false discovery rate (FDR) and recall rate. Interestingly, a few scRNA-seq DEMs obtained results comparable with the best-performing bulk RNA-seq tools. Almost all DEMs' performance improved when increasing the number of replicates. CircRNA expression studies require careful design, choice of DEM and DEM configuration. This analysis can guide scientists in selecting the appropriate tools to investigate circRNA differential expression with RNA-seq experiments.

Combination therapy is a promising strategy for confronting the complexity of cancer. However, experimental exploration of the vast space of potential drug combinations is costly and unfeasible. Therefore, computational methods for predicting drug synergy are much needed for narrowing down this space, especially when examining new cellular contexts. Here, we thus introduce CCSynergy, a flexible, context aware and integrative deep-learning framework that we have established to unleash the potential of the Chemical Checker extended drug bioactivity profiles for the purpose of drug synergy prediction. We have shown that CCSynergy enables predictions of superior accuracy, remarkable robustness and improved context generalizability as compared to the state-of-the-art methods in the field. Having established the potential of CCSynergy for generating experimentally validated predictions, we next exhaustively explored the untested drug combination space. This resulted in a compendium of potentially synergistic drug combinations on hundreds of cancer cell lines, which can guide future experimental screens.

BACKGROUND: Cell-penetrating peptides (CPPs) have received considerable attention as a means of transporting pharmacologically active molecules into living cells without damaging the cell membrane, and thus hold great promise as future therapeutics. Recently, several machine learning-based algorithms have been proposed for predicting CPPs. However, most existing predictive methods do not consider the agreement (disagreement) between similar (dissimilar) CPPs and depend heavily on expert knowledge-based handcrafted features. RESULTS: In this study, we present SiameseCPP, a novel deep learning framework for automated CPPs prediction. SiameseCPP learns discriminative representations of CPPs based on a well-pretrained model and a Siamese neural network consisting of a transformer and gated recurrent units. Contrastive learning is used for the first time to build a CPP predictive model. Comprehensive experiments demonstrate that our proposed SiameseCPP is superior to existing baseline models for predicting CPPs. Moreover, SiameseCPP also achieves good performance on other functional peptide datasets, exhibiting satisfactory generalization ability.

Long noncoding ribonucleic acids (RNAs; LncRNAs) endowed with both protein-coding and noncoding functions are referred to as 'dual functional lncRNAs'. Recently, dual functional lncRNAs have been intensively studied and identified as involved in various fundamental cellular processes. However, apart from time-consuming and cell-type-specific experiments, there is virtually no in silico method for predicting the identity of dual functional lncRNAs. Here, we developed a deep-learning model with a multi-head self-attention mechanism, LncReader, to identify dual functional lncRNAs. Our data demonstrated that LncReader showed multiple advantages compared to various classical machine learning methods using benchmark datasets from our previously reported cncRNAdb project. Moreover, to obtain independent in-house datasets for robust testing, mass spectrometry proteomics combined with RNA-seq and Ribo-seq were applied in four leukaemia cell lines, which further confirmed that LncReader achieved the best performance compared to other tools. Therefore, LncReader provides an accurate and practical tool that enables fast dual functional lncRNA identification.

Protein language modeling is a fast-emerging deep learning method in bioinformatics with diverse applications such as structure prediction and protein design. However, application toward estimating sequence conservation for functional site prediction has not been systematically explored. Here, we present a method for the alignment-free estimation of sequence conservation using sequence embeddings generated from protein language models. Comprehensive benchmarks across publicly available protein language models reveal that ESM2 models provide the best performance to computational cost ratio for conservation estimation. Applying our method to full-length protein sequences, we demonstrate that embedding-based methods are not sensitive to the order of conserved elements-conservation scores can be calculated for multidomain proteins in a single run, without the need to separate individual domains. Our method can also identify conserved functional sites within fast-evolving sequence regions (such as domain inserts), which we demonstrate through the identification of conserved phosphorylation motifs in variable insert segments in protein kinases. Overall, embedding-based conservation analysis is a broadly applicable method for identifying potential functional sites in any full-length protein sequence and estimating conservation in an alignment-free manner. To run this on your protein sequence of interest, try our scripts at https://github.com/esbgkannan/kibby.

Circular RNAs (circRNAs) are reverse-spliced and covalently closed RNAs. Their interactions with RNA-binding proteins (RBPs) have multiple effects on the progress of many diseases. Some computational methods are proposed to identify RBP binding sites on circRNAs but suffer from insufficient accuracy, robustness and explanation. In this study, we first take the characteristics of both RNA and RBP into consideration. We propose a method for discriminating circRNA-RBP binding sites based on multi-scale characterizing sequence and structure features, called CRMSS. For circRNAs, we use sequence ${k}\hbox{-}{mer}$ embedding and the forming probabilities of local secondary structures as features. For RBPs, we combine sequence and structure frequencies of RNA-binding domain regions to generate features. We capture binding patterns with multi-scale residual blocks. With BiLSTM and attention mechanism, we obtain the contextual information of high-level representation for circRNA-RBP binding. To validate the effectiveness of CRMSS, we compare its predictive performance with other methods on 37 RBPs. Taking the properties of both circRNAs and RBPs into account, CRMSS achieves superior performance over state-of-the-art methods. In the case study, our model provides reliable predictions and correctly identifies experimentally verified circRNA-RBP pairs. The code of CRMSS is freely available at https://github.com/BioinformaticsCSU/CRMSS.

The progress of single-cell RNA sequencing (scRNA-seq) has led to a large number of scRNA-seq data, which are widely used in biomedical research. The noise in the raw data and tens of thousands of genes pose a challenge to capture the real structure and effective information of scRNA-seq data. Most of the existing single-cell analysis methods assume that the low-dimensional embedding of the raw data belongs to a Gaussian distribution or a low-dimensional nonlinear space without any prior information, which limits the flexibility and controllability of the model to a great extent. In addition, many existing methods need high computational cost, which makes them difficult to be used to deal with large-scale datasets. Here, we design and develop a depth generation model named Gaussian mixture adversarial autoencoders (scGMAAE), assuming that the low-dimensional embedding of different types of cells follows different Gaussian distributions, integrating Bayesian variational inference and adversarial training, as to give the interpretable latent representation of complex data and discover the statistical distribution of different types of cells. The scGMAAE is provided with good controllability, interpretability and scalability. Therefore, it can process large-scale datasets in a short time and give competitive results. scGMAAE outperforms existing methods in several ways, including dimensionality reduction visualization, cell clustering, differential expression analysis and batch effect removal. Importantly, compared with most deep learning methods, scGMAAE requires less iterations to generate the best results.

Protein-protein interactions (PPIs) are a major component of the cellular biochemical reaction network. Rich sequence information and machine learning techniques reduce the dependence of exploring PPIs on wet experiments, which are costly and time-consuming. This paper proposes a PPI prediction model, multi-scale architecture residual network for PPIs (MARPPI), based on dual-channel and multi-feature. Multi-feature leverages Res2vec to obtain the association information between residues, and utilizes pseudo amino acid composition, autocorrelation descriptors and multivariate mutual information to achieve the amino acid composition and order information, physicochemical properties and information entropy, respectively. Dual channel utilizes multi-scale architecture improved ResNet network which extracts protein sequence features to reduce protein feature loss. Compared with other advanced methods, MARPPI achieves 96.03%, 99.01% and 91.80% accuracy in the intraspecific datasets of Saccharomyces cerevisiae, Human and Helicobacter pylori, respectively. The accuracy on the two interspecific datasets of Human-Bacillus anthracis and Human-Yersinia pestis is 97.29%, and 95.30%, respectively. In addition, results on specific datasets of disease (neurodegenerative and metabolic disorders) demonstrate the ability to detect hidden interactions. To better illustrate the performance of MARPPI, evaluations on independent datasets and PPIs network suggest that MARPPI can be used to predict cross-species interactions. The above shows that MARPPI can be regarded as a concise, efficient and accurate tool for PPI datasets.

Machine learning-based scoring functions (MLSFs) have become a very favorable alternative to classical scoring functions because of their potential superior screening performance. However, the information of negative data used to construct MLSFs was rarely reported in the literature, and meanwhile the putative inactive molecules recorded in existing databases usually have obvious bias from active molecules. Here we proposed an easy-to-use method named AMLSF that combines active learning using negative molecular selection strategies with MLSF, which can iteratively improve the quality of inactive sets and thus reduce the false positive rate of virtual screening. We chose energy auxiliary terms learning as the MLSF and validated our method on eight targets in the diverse subset of DUD-E. For each target, we screened the IterBioScreen database by AMLSF and compared the screening results with those of the four control models. The results illustrate that the number of active molecules in the top 1000 molecules identified by AMLSF was significantly higher than those identified by the control models. In addition, the free energy calculation results for the top 10 molecules screened out by the AMLSF, null model and control models based on DUD-E also proved that more active molecules can be identified, and the false positive rate can be reduced by AMLSF.

Type 1 diabetes (T1D) outcome prediction plays a vital role in identifying novel risk factors, ensuring early patient care and designing cohort studies. TEDDY is a longitudinal cohort study that collects a vast amount of multi-omics and clinical data from its participants to explore the progression and markers of T1D. However, missing data in the omics profiles make the outcome prediction a difficult task. TEDDY collected time series gene expression for less than 6% of enrolled participants. Additionally, for the participants whose gene expressions are collected, 79% time steps are missing. This study introduces an advanced bioinformatics framework for gene expression imputation and islet autoimmunity (IA) prediction. The imputation model generates synthetic data for participants with partially or entirely missing gene expression. The prediction model integrates the synthetic gene expression with other risk factors to achieve better predictive performance. Comprehensive experiments on TEDDY datasets show that: (1) Our pipeline can effectively integrate synthetic gene expression with family history, HLA genotype and SNPs to better predict IA status at 2 years (sensitivity 0.622, AUC 0.715) compared with the individual datasets and state-of-the-art results in the literature (AUC 0.682). (2) The synthetic gene expression contains predictive signals as strong as the true gene expression, reducing reliance on expensive and long-term longitudinal data collection. (3) Time series gene expression is crucial to the proposed improvement and shows significantly better predictive ability than cross-sectional gene expression. (4) Our pipeline is robust to limited data availability. Availability: Code is available at https://github.com/compbiolabucf/TEDDY.

Recent developments of deep learning methods have demonstrated their feasibility in liver malignancy diagnosis using ultrasound (US) images. However, most of these methods require manual selection and annotation of US images by radiologists, which limit their practical application. On the other hand, US videos provide more comprehensive morphological information about liver masses and their relationships with surrounding structures than US images, potentially leading to a more accurate diagnosis. Here, we developed a fully automated artificial intelligence (AI) pipeline to imitate the workflow of radiologists for detecting liver masses and diagnosing liver malignancy. In this pipeline, we designed an automated mass-guided strategy that used segmentation information to direct diagnostic models to focus on liver masses, thus increasing diagnostic accuracy. The diagnostic models based on US videos utilized bi-directional convolutional long short-term memory modules with an attention-boosted module to learn and fuse spatiotemporal information from consecutive video frames. Using a large-scale dataset of 50 063 US images and video frames from 11 468 patients, we developed and tested the AI pipeline and investigated its applications. A dataset of annotated US images is available at https://doi.org/10.5281/zenodo.7272660.

Protein-ligand binding affinity prediction is an important task in structural bioinformatics for drug discovery and design. Although various scoring functions (SFs) have been proposed, it remains challenging to accurately evaluate the binding affinity of a protein-ligand complex with the known bound structure because of the potential preference of scoring system. In recent years, deep learning (DL) techniques have been applied to SFs without sophisticated feature engineering. Nevertheless, existing methods cannot model the differential contribution of atoms in various regions of proteins, and the relationship between atom properties and intermolecular distance is also not fully explored. We propose a novel empirical graph neural network for accurate protein-ligand binding affinity prediction (EGNA). Graphs of protein, ligand and their interactions are constructed based on different regions of each bound complex. Proteins and ligands are effectively represented by graph convolutional layers, enabling the EGNA to capture interaction patterns precisely by simulating empirical SFs. The contributions of different factors on binding affinity can thus be transparently investigated. EGNA is compared with the state-of-the-art machine learning-based SFs on two widely used benchmark data sets. The results demonstrate the superiority of EGNA and its good generalization capability.

Understanding drug selectivity mechanism is a long-standing issue for helping design drugs with high specificity. Designing drugs targeting cyclin-dependent kinases (CDKs) with high selectivity is challenging because of their highly conserved binding pockets. To reveal the underlying general selectivity mechanism, we carried out comprehensive analyses from both the thermodynamics and kinetics points of view on a representative CDK12 inhibitor. To fully capture the binding features of the drug-target recognition process, we proposed to use kinetic residue energy analysis (KREA) in conjunction with the community network analysis (CNA) to reveal the underlying cooperation effect between individual residues/protein motifs to the binding/dissociating process of the ligand. The general mechanism of drug selectivity in CDKs can be summarized as that the difference of structural cooperation between the ligand and the protein motifs leads to the difference of the energetic contribution of the key residues to the ligand. The proposed mechanisms may be prevalent in drug selectivity issues, and the insights may help design new strategies to overcome/attenuate the drug selectivity associated problems.

Recent studies have shown that the expression of circRNAs would affect drug sensitivity of cells and thus significantly influence the efficacy of drugs. Traditional biomedical experiments to validate such relationships are time-consuming and costly. Therefore, developing effective computational methods to predict potential associations between circRNAs and drug sensitivity is an important and urgent task. In this study, we propose a novel method, called MNGACDA, to predict possible circRNA-drug sensitivity associations for further biomedical screening. First, MNGACDA uses multiple sources of information from circRNAs and drugs to construct multimodal networks. It then employs node-level attention graph auto-encoders to obtain low-dimensional embeddings for circRNAs and drugs from the multimodal networks. Finally, an inner product decoder is applied to predict the association scores between circRNAs and drug sensitivity based on the embedding representations of circRNAs and drugs. Extensive experimental results based on cross-validations show that MNGACDA outperforms six other state-of-the-art methods. Furthermore, excellent performance in case studies demonstrates that MNGACDA is an effective tool for predicting circRNA-drug sensitivity associations in real situations. These results confirm the reliable prediction ability of MNGACDA in revealing circRNA-drug sensitivity associations.

Drug-drug interaction (DDI) prediction identifies interactions of drug combinations in which the adverse side effects caused by the physicochemical incompatibility have attracted much attention. Previous studies usually model drug information from single or dual views of the whole drug molecules but ignore the detailed interactions among atoms, which leads to incomplete and noisy information and limits the accuracy of DDI prediction. In this work, we propose a novel dual-view drug representation learning network for DDI prediction ('DSN-DDI'), which employs local and global representation learning modules iteratively and learns drug substructures from the single drug ('intra-view') and the drug pair ('inter-view') simultaneously. Comprehensive evaluations demonstrate that DSN-DDI significantly improved performance on DDI prediction for the existing drugs by achieving a relatively improved accuracy of 13.01% and an over 99% accuracy under the transductive setting. More importantly, DSN-DDI achieves a relatively improved accuracy of 7.07% to unseen drugs and shows the usefulness for real-world DDI applications. Finally, DSN-DDI exhibits good transferability on synergistic drug combination prediction and thus can serve as a generalized framework in the drug discovery field.

The inference of gene regulatory networks (GRNs) is of great importance for understanding the complex regulatory mechanisms within cells. The emergence of single-cell RNA-sequencing (scRNA-seq) technologies enables the measure of gene expression levels for individual cells, which promotes the reconstruction of GRNs at single-cell resolution. However, existing network inference methods are mainly designed for data collected from a single data source, which ignores the information provided by multiple related data sources. In this paper, we propose a multi-view contrastive learning (DeepMCL) model to infer GRNs from scRNA-seq data collected from multiple data sources or time points. We first represent each gene pair as a set of histogram images, and then introduce a deep Siamese convolutional neural network with contrastive loss to learn the low-dimensional embedding for each gene pair. Moreover, an attention mechanism is introduced to integrate the embeddings extracted from different data sources and different neighbor gene pairs. Experimental results on synthetic and real-world datasets validate the effectiveness of our contrastive learning and attention mechanisms, demonstrating the effectiveness of our model in integrating multiple data sources for GRN inference.

Machine learning including modern deep learning models has been extensively used in drug design and screening. However, reliable prediction of molecular properties is still challenging when exploring out-of-domain regimes, even for deep neural networks. Therefore, it is important to understand the uncertainty of model predictions, especially when the predictions are used to guide further experiments. In this study, we explored the utility and effectiveness of evidential uncertainty in compound screening. The evidential Graphormer model was proposed for uncertainty-guided discovery of KDM1A/LSD1 inhibitors. The benchmarking results illustrated that (i) Graphormer exhibited comparative predictive power to state-of-the-art models, and (ii) evidential regression enabled well-ranked uncertainty estimates and calibrated predictions. Subsequently, we leveraged time-splitting on the curated KDM1A/LSD1 dataset to simulate out-of-distribution predictions. The retrospective virtual screening showed that the evidential uncertainties helped reduce false positives among the top-acquired compounds and thus enabled higher experimental validation rates. The trained model was then used to virtually screen an independent in-house compound set. The top 50 compounds ranked by two different ranking strategies were experimentally validated, respectively. In general, our study highlighted the importance to understand the uncertainty in prediction, which can be recognized as an interpretable dimension to model predictions.

Graph neural networks based on deep learning methods have been extensively applied to the molecular property prediction because of its powerful feature learning ability and good performance. However, most of them are black boxes and cannot give the reasonable explanation about the underlying prediction mechanisms, which seriously reduce people's trust on the neural network-based prediction models. Here we proposed a novel graph neural network named iteratively focused graph network (IFGN), which can gradually identify the key atoms/groups in the molecule that are closely related to the predicted properties by the multistep focus mechanism. At the same time, the combination of the multistep focus mechanism with visualization can also generate multistep interpretations, thus allowing us to gain a deep understanding of the predictive behaviors of the model. For all studied eight datasets, the IFGN model achieved good prediction performance, indicating that the proposed multistep focus mechanism also can improve the performance of the model obviously besides increasing the interpretability of built model. For researchers to use conveniently, the corresponding website (http://graphadmet.cn/works/IFGN) was also developed and can be used free of charge.

Protein model quality assessment plays an important role in protein structure prediction, protein design and drug discovery. In this work, DeepUMQA2, a substantially improved version of DeepUMQA for protein model quality assessment, is proposed. First, sequence features containing protein co-evolution information and structural features reflecting family information are extracted to complement model-dependent features. Second, a novel backbone network based on triangular multiplication update and axial attention mechanism is designed to enhance information exchange between inter-residue pairs. On CASP13 and CASP14 datasets, the performance of DeepUMQA2 increases by 20.5 and 20.4% compared with DeepUMQA, respectively (measured by top 1 loss). Moreover, on the three-month CAMEO dataset (11 March to 04 June 2022), DeepUMQA2 outperforms DeepUMQA by 15.5% (measured by local AUC0,0.2) and ranks first among all competing server methods in CAMEO blind test. Experimental results show that DeepUMQA2 outperforms state-of-the-art model quality assessment methods, such as ProQ3D-LDDT, ModFOLD8, and DeepAccNet and DeepUMQA2 can select more suitable best models than state-of-the-art protein structure methods, such as AlphaFold2, RoseTTAFold and I-TASSER, provided themselves.

The prediction of peptide and protein function is important for research and industrial applications, and many machine learning methods have been developed for this purpose. The existing models have encountered many challenges, including the lack of effective and comprehensive features and the limited applicability of each model. Here, we introduce an Integrated Peptide and Protein function prediction Framework based on Fused features and Ensemble models (IPPF-FE), which can accurately capture the relationship between features and labels. The results indicated that IPPF-FE outperformed existing state-of-the-art (SOTA) models on more than 8 different categories of peptide and protein tasks. In addition, t-distributed Stochastic Neighbour Embedding demonstrated the advantages of IPPF-FE. We anticipate that our method will become a versatile tool for peptide and protein prediction tasks and shed light on the future development of related models. The model is open source and available in the GitHub repository https://github.com/Luo-SynBioLab/IPPF-FE.

Long noncoding RNA (lncRNA) is a kind of noncoding RNA with a length of more than 200 nucleotide units. Numerous research studies have proven that although lncRNAs cannot be directly translated into proteins, lncRNAs still play an important role in human growth processes by interacting with proteins. Since traditional biological experiments often require a lot of time and material costs to explore potential lncRNA-protein interactions (LPI), several computational models have been proposed for this task. In this study, we introduce a novel deep learning method known as combined graph auto-encoders (LPICGAE) to predict potential human LPIs. First, we apply a variational graph auto-encoder to learn the low dimensional representations from the high-dimensional features of lncRNAs and proteins. Then the graph auto-encoder is used to reconstruct the adjacency matrix for inferring potential interactions between lncRNAs and proteins. Finally, we minimize the loss of the two processes alternately to gain the final predicted interaction matrix. The result in 5-fold cross-validation experiments illustrates that our method achieves an average area under receiver operating characteristic curve of 0.974 and an average accuracy of 0.985, which is better than those of existing six state-of-the-art computational methods. We believe that LPICGAE can help researchers to gain more potential relationships between lncRNAs and proteins effectively.

Effective full quantum mechanics (FQM) calculation of protein remains a grand challenge and of great interest in computational biology with substantial applications in drug discovery, protein dynamic simulation and protein folding. However, the huge computational complexity of the existing QM methods impends their applications in large systems. Here, we design a transfer-learning-based deep learning (TDL) protocol for effective FQM calculations (TDL-FQM) on proteins. By incorporating a transfer-learning algorithm into deep neural network (DNN), the TDL-FQM protocol is capable of performing calculations at any given accuracy using models trained from small datasets with high-precision and knowledge learned from large amount of low-level calculations. The high-level double-hybrid DFT functional and high-level quality of basis set is used in this work as a case study to evaluate the performance of TDL-FQM, where the selected 15 proteins are predicted to have a mean absolute error of 0.01 kcal/mol/atom for potential energy and an average root mean square error of 1.47 kcal/mol/$ {\rm A^{^{ \!\!\!o}}} $ for atomic forces. The proposed TDL-FQM approach accelerates the FQM calculation more than thirty thousand times faster in average and presents more significant benefits in efficiency as the size of protein increases. The ability to learn knowledge from one task to solve related problems demonstrates that the proposed TDL-FQM overcomes the limitation of standard DNN and has a strong power to predict proteins with high precision, which solves the challenge of high precision prediction in large chemical and biological systems.

Accounting for cell type compositions has been very successful at analyzing high-throughput data from heterogeneous tissues. Differential gene expression analysis at cell type level is becoming increasingly popular, yielding biomarker discovery in a finer granularity within a particular cell type. Although several computational methods have been developed to identify cell type-specific differentially expressed genes (csDEG) from RNA-seq data, a systematic evaluation is yet to be performed. Here, we thoroughly benchmark six recently published methods: CellDMC, CARseq, TOAST, LRCDE, CeDAR and TCA, together with two classical methods, csSAM and DESeq2, for a comprehensive comparison. We aim to systematically evaluate the performance of popular csDEG detection methods and provide guidance to researchers. In simulation studies, we benchmark available methods under various scenarios of baseline expression levels, sample sizes, cell type compositions, expression level alterations, technical noises and biological dispersions. Real data analyses of three large datasets on inflammatory bowel disease, lung cancer and autism provide evaluation in both the gene level and the pathway level. We find that csDEG calling is strongly affected by effect size, baseline expression level and cell type compositions. Results imply that csDEG discovery is a challenging task itself, with room to improvements on handling low signal-to-noise ratio and low expression genes.

Emerging evidence has proved that circular RNAs (circRNAs) are implicated in pathogenic processes. They are regarded as promising biomarkers for diagnosis due to covalently closed loop structures. As opposed to traditional experiments, computational approaches can identify circRNA-disease associations at a lower cost. Aggregating multi-source pathogenesis data helps to alleviate data sparsity and infer potential associations at the system level. The majority of computational approaches construct a homologous network using multi-source data, but they lose the heterogeneity of the data. Effective methods that use the features of multi-source data are considered as a matter of urgency. In this paper, we propose a model (CDHGNN) based on edge-weighted graph attention and heterogeneous graph neural networks for potential circRNA-disease association prediction. The circRNA network, micro RNA network, disease network and heterogeneous network are constructed based on multi-source data. To reflect association probabilities between nodes, an edge-weighted graph attention network model is designed for node features. To assign attention weights to different types of edges and learn contextual meta-path, CDHGNN infers potential circRNA-disease association based on heterogeneous neural networks. CDHGNN outperforms state-of-the-art algorithms in terms of accuracy. Edge-weighted graph attention networks and heterogeneous graph networks have both improved performance significantly. Furthermore, case studies suggest that CDHGNN is capable of identifying specific molecular associations and investigating biomolecular regulatory relationships in pathogenesis. The code of CDHGNN is freely available at https://github.com/BioinformaticsCSU/CDHGNN.

MicroRNA (miRNA)-target interaction (MTI) plays a substantial role in various cell activities, molecular regulations and physiological processes. Published biomedical literature is the carrier of high-confidence MTI knowledge. However, digging out this knowledge in an efficient manner from large-scale published articles remains challenging. To address this issue, we were motivated to construct a deep learning-based model. We applied the pre-trained language models to biomedical text to obtain the representation, and subsequently fed them into a deep neural network with gate mechanism layers and a fully connected layer for the extraction of MTI information sentences. Performances of the proposed models were evaluated using two datasets constructed on the basis of text data obtained from miRTarBase. The validation and test results revealed that incorporating both PubMedBERT and SciBERT for sentence level encoding with the long short-term memory (LSTM)-based deep neural network can yield an outstanding performance, with both F1 and accuracy being higher than 80% on validation data and test data. Additionally, the proposed deep learning method outperformed the following machine learning methods: random forest, support vector machine, logistic regression and bidirectional LSTM. This work would greatly facilitate studies on MTI analysis and regulations. It is anticipated that this work can assist in large-scale screening of miRNAs, thereby revealing their functional roles in various diseases, which is important for the development of highly specific drugs with fewer side effects. Source code and corpus are publicly available at https://github.com/qi29.

Monoclonal antibodies are biotechnologically produced proteins with various applications in research, therapeutics and diagnostics. Their ability to recognize and bind to specific molecule structures makes them essential research tools and therapeutic agents. Sequence information of antibodies is helpful for understanding antibody-antigen interactions and ensuring their affinity and specificity. De novo protein sequencing based on mass spectrometry is a valuable method to obtain the amino acid sequence of peptides and proteins without a priori knowledge. In this study, we evaluated six recently developed de novo peptide sequencing algorithms (Novor, pNovo 3, DeepNovo, SMSNet, PointNovo and Casanovo), which were not specifically designed for antibody data. We validated their ability to identify and assemble antibody sequences on three multi-enzymatic data sets. The deep learning-based tools Casanovo and PointNovo showed an increased peptide recall across different enzymes and data sets compared with spectrum-graph-based approaches. We evaluated different error types of de novo peptide sequencing tools and their performance for different numbers of missing cleavage sites, noisy spectra and peptides of various lengths. We achieved a sequence coverage of 97.69-99.53% on the light chains of three different antibody data sets using the de Bruijn assembler ALPS and the predictions from Casanovo. However, low sequence coverage and accuracy on the heavy chains demonstrate that complete de novo protein sequencing remains a challenging issue in proteomics that requires improved de novo error correction, alternative digestion strategies and hybrid approaches such as homology search to achieve high accuracy on long protein sequences.

The accurate prediction of cancer drug sensitivity according to the multiomics profiles of individual patients is crucial for precision cancer medicine. However, the development of prediction models has been challenged by the complex crosstalk of input features and the resistance-dominant drug response information contained in public databases. In this study, we propose a novel multidrug response prediction framework, response-aware multitask prediction (RAMP), via a Bayesian neural network and restrict it by soft-supervised contrastive regularization. To utilize network embedding vectors as representation learning features for heterogeneous networks, we harness response-aware negative sampling, which applies cell line-drug response information to the training of network embeddings. RAMP overcomes the prediction accuracy limitation induced by the imbalance of trained response data based on the comprehensive selection and utilization of drug response features. When trained on the Genomics of Drug Sensitivity in Cancer dataset, RAMP achieved an area under the receiver operating characteristic curve > 89%, an area under the precision-recall curve > 59% and an $\textrm{F}_1$ score > 52% and outperformed previously developed methods on both balanced and imbalanced datasets. Furthermore, RAMP predicted many missing drug responses that were not included in the public databases. Our results showed that RAMP will be suitable for the high-throughput prediction of cancer drug sensitivity and will be useful for guiding cancer drug selection processes. The Python implementation for RAMP is available at https://github.com/hvcl/RAMP.

The gut microbial communities are highly plastic throughout life, and the human gut microbial communities show spatial-temporal dynamic patterns at different life stages. However, the underlying association between gut microbial communities and time-related factors remains unclear. The lack of context-awareness, insufficient data, and the existence of batch effect are the three major issues, making the life trajection of the host based on gut microbial communities problematic. Here, we used a novel computational approach (microDELTA, microbial-based deep life trajectory) to track longitudinal human gut microbial communities' alterations, which employs transfer learning for context-aware mining of gut microbial community dynamics at different life stages. Using an infant cohort, we demonstrated that microDELTA outperformed Neural Network for accurately predicting the age of infant with different delivery mode, especially for newborn infants of vaginal delivery with the area under the receiver operating characteristic curve of microDELTA and Neural Network at 0.811 and 0.436, respectively. In this context, we have discovered the influence of delivery mode on infant gut microbial communities. Along the human lifespan, we also applied microDELTA to a Chinese traveler cohort, a Hadza hunter-gatherer cohort and an elderly cohort. Results revealed the association between long-term dietary shifts during travel and adult gut microbial communities, the seasonal cycling of gut microbial communities for the Hadza hunter-gatherers, and the distinctive microbial pattern of elderly gut microbial communities. In summary, microDELTA can largely solve the issues in tracing the life trajectory of the human microbial communities and generate accurate and flexible models for a broad spectrum of microbial-based longitudinal researches.

The advances in single-cell ribonucleic acid sequencing (scRNA-seq) allow researchers to explore cellular heterogeneity and human diseases at cell resolution. Cell clustering is a prerequisite in scRNA-seq analysis since it can recognize cell identities. However, the high dimensionality, noises and significant sparsity of scRNA-seq data have made it a big challenge. Although many methods have emerged, they still fail to fully explore the intrinsic properties of cells and the relationship among cells, which seriously affects the downstream clustering performance. Here, we propose a new deep contrastive clustering algorithm called scDCCA. It integrates a denoising auto-encoder and a dual contrastive learning module into a deep clustering framework to extract valuable features and realize cell clustering. Specifically, to better characterize and learn data representations robustly, scDCCA utilizes a denoising Zero-Inflated Negative Binomial model-based auto-encoder to extract low-dimensional features. Meanwhile, scDCCA incorporates a dual contrastive learning module to capture the pairwise proximity of cells. By increasing the similarities between positive pairs and the differences between negative ones, the contrasts at both the instance and the cluster level help the model learn more discriminative features and achieve better cell segregation. Furthermore, scDCCA joins feature learning with clustering, which realizes representation learning and cell clustering in an end-to-end manner. Experimental results of 14 real datasets validate that scDCCA outperforms eight state-of-the-art methods in terms of accuracy, generalizability, scalability and efficiency. Cell visualization and biological analysis demonstrate that scDCCA significantly improves clustering and facilitates downstream analysis for scRNA-seq data. The code is available at https://github.com/WJ319/scDCCA.

Underlying medical conditions, such as cancer, kidney disease and heart failure, are associated with a higher risk for severe COVID-19. Accurate classification of COVID-19 patients with underlying medical conditions is critical for personalized treatment decision and prognosis estimation. In this study, we propose an interpretable artificial intelligence model termed VDJMiner to mine the underlying medical conditions and predict the prognosis of COVID-19 patients according to their immune repertoires. In a cohort of more than 1400 COVID-19 patients, VDJMiner accurately identifies multiple underlying medical conditions, including cancers, chronic kidney disease, autoimmune disease, diabetes, congestive heart failure, coronary artery disease, asthma and chronic obstructive pulmonary disease, with an average area under the receiver operating characteristic curve (AUC) of 0.961. Meanwhile, in this same cohort, VDJMiner achieves an AUC of 0.922 in predicting severe COVID-19. Moreover, VDJMiner achieves an accuracy of 0.857 in predicting the response of COVID-19 patients to tocilizumab treatment on the leave-one-out test. Additionally, VDJMiner interpretively mines and scores V(D)J gene segments of the T-cell receptors that are associated with the disease. The identified associations between single-cell V(D)J gene segments and COVID-19 are highly consistent with previous studies. The source code of VDJMiner is publicly accessible at https://github.com/TencentAILabHealthcare/VDJMiner. The web server of VDJMiner is available at https://gene.ai.tencent.com/VDJMiner/.

Accurate variant pathogenicity predictions are important in genetic studies of human diseases. Inframe insertion and deletion variants (indels) alter protein sequence and length, but not as deleterious as frameshift indels. Inframe indel Interpretation is challenging due to limitations in the available number of known pathogenic variants for training. Existing prediction methods largely use manually encoded features including conservation, protein structure and function, and allele frequency to infer variant pathogenicity. Recent advances in deep learning modeling of protein sequences and structures provide an opportunity to improve the representation of salient features based on large numbers of protein sequences. We developed a new pathogenicity predictor for SHort Inframe iNsertion and dEletion (SHINE). SHINE uses pretrained protein language models to construct a latent representation of an indel and its protein context from protein sequences and multiple protein sequence alignments, and feeds the latent representation into supervised machine learning models for pathogenicity prediction. We curated training data from ClinVar and gnomAD, and created two test datasets from different sources. SHINE achieved better prediction performance than existing methods for both deletion and insertion variants in these two test datasets. Our work suggests that unsupervised protein language models can provide valuable information about proteins, and new methods based on these models can improve variant interpretation in genetic analyses.

DNA-binding proteins (DBPs) play crucial roles in numerous cellular processes including nucleotide recognition, transcriptional control and the regulation of gene expression. Majority of the existing computational techniques for identifying DBPs are mainly applicable to human and mouse datasets. Even though some models have been tested on Arabidopsis, they produce poor accuracy when applied to other plant species. Therefore, it is imperative to develop an effective computational model for predicting plant DBPs. In this study, we developed a comprehensive computational model for plant specific DBPs identification. Five shallow learning and six deep learning models were initially used for prediction, where shallow learning methods outperformed deep learning algorithms. In particular, support vector machine achieved highest repeated 5-fold cross-validation accuracy of 94.0% area under receiver operating characteristic curve (AUC-ROC) and 93.5% area under precision recall curve (AUC-PR). With an independent dataset, the developed approach secured 93.8% AUC-ROC and 94.6% AUC-PR. While compared with the state-of-art existing tools by using an independent dataset, the proposed model achieved much higher accuracy. Overall results suggest that the developed computational model is more efficient and reliable as compared to the existing models for the prediction of DBPs in plants. For the convenience of the majority of experimental scientists, the developed prediction server PlDBPred is publicly accessible at https://iasri-sg.icar.gov.in/pldbpred/.The source code is also provided at https://iasri-sg.icar.gov.in/pldbpred/source_code.php for prediction using a large-size dataset.

Due to its promising capacity in improving drug efficacy, polypharmacology has emerged to be a new theme in the drug discovery of complex disease. In the process of novel multi-target drugs (MTDs) discovery, in silico strategies come to be quite essential for the advantage of high throughput and low cost. However, current researchers mostly aim at typical closely related target pairs. Because of the intricate pathogenesis networks of complex diseases, many distantly related targets are found to play crucial role in synergistic treatment. Therefore, an innovational method to develop drugs which could simultaneously target distantly related target pairs is of utmost importance. At the same time, reducing the false discovery rate in the design of MTDs remains to be the daunting technological difficulty. In this research, effective small molecule clustering in the positive dataset, together with a putative negative dataset generation strategy, was adopted in the process of model constructions. Through comprehensive assessment on 10 target pairs with hierarchical similarity-levels, the proposed strategy turned out to reduce the false discovery rate successfully. Constructed model types with much smaller numbers of inhibitor molecules gained considerable yields and showed better false-hit controllability than before. To further evaluate the generalization ability, an in-depth assessment of high-throughput virtual screening on ChEMBL database was conducted. As a result, this novel strategy could hierarchically improve the enrichment factors for each target pair (especially for those distantly related/unrelated target pairs), corresponding to target pair similarity-levels.

Drug combinations could trigger pharmacological therapeutic effects (TEs) and adverse effects (AEs). Many computational methods have been developed to predict TEs, e.g. the therapeutic synergy scores of anti-cancer drug combinations, or AEs from drug-drug interactions. However, most of the methods treated the AEs and TEs predictions as two separate tasks, ignoring the potential mechanistic commonalities shared between them. Based on previous clinical observations, we hypothesized that by learning the shared mechanistic commonalities between AEs and TEs, we could learn the underlying MoAs (mechanisms of actions) and ultimately improve the accuracy of TE predictions. To test our hypothesis, we formulated the TE prediction problem as a multi-task heterogeneous network learning problem that performed TE and AE learning tasks simultaneously. To solve this problem, we proposed Muthene (multi-task heterogeneous network embedding) and evaluated it on our collected drug-drug interaction dataset with both TEs and AEs indications. Our experimental results showed that, by including the AE prediction as an auxiliary task, Muthene generated more accurate TE predictions than standard single-task learning methods, which supports our hypothesis. Using a drug pair Vincristine-Dasatinib as a case study, we demonstrated that our method not only provides a novel way of TE predictions but also helps us gain a deeper understanding of the MoAs of drug combinations.

Driven by multi-omics data, some multi-view clustering algorithms have been successfully applied to cancer subtypes prediction, aiming to identify subtypes with biometric differences in the same cancer, thereby improving the clinical prognosis of patients and designing personalized treatment plan. Due to the fact that the number of patients in omics data is much smaller than the number of genes, multi-view spectral clustering based on similarity learning has been widely developed. However, these algorithms still suffer some problems, such as over-reliance on the quality of pre-defined similarity matrices for clustering results, inability to reasonably handle noise and redundant information in high-dimensional omics data, ignoring complementary information between omics data, etc. This paper proposes multi-view spectral clustering with latent representation learning (MSCLRL) method to alleviate the above problems. First, MSCLRL generates a corresponding low-dimensional latent representation for each omics data, which can effectively retain the unique information of each omics and improve the robustness and accuracy of the similarity matrix. Second, the obtained latent representations are assigned appropriate weights by MSCLRL, and global similarity learning is performed to generate an integrated similarity matrix. Third, the integrated similarity matrix is used to feed back and update the low-dimensional representation of each omics. Finally, the final integrated similarity matrix is used for clustering. In 10 benchmark multi-omics datasets and 2 separate cancer case studies, the experiments confirmed that the proposed method obtained statistically and biologically meaningful cancer subtypes.

Current machine learning-based methods have achieved inspiring predictions in the scenarios of mono-type and multi-type drug-drug interactions (DDIs), but they all ignore enhancive and depressive pharmacological changes triggered by DDIs. In addition, these pharmacological changes are asymmetric since the roles of two drugs in an interaction are different. More importantly, these pharmacological changes imply significant topological patterns among DDIs. To address the above issues, we first leverage Balance theory and Status theory in social networks to reveal the topological patterns among directed pharmacological DDIs, which are modeled as a signed and directed network. Then, we design a novel graph representation learning model named SGRL-DDI (social theory-enhanced graph representation learning for DDI) to realize the multitask prediction of DDIs. SGRL-DDI model can capture the task-joint information by integrating relation graph convolutional networks with Balance and Status patterns. Moreover, we utilize task-specific deep neural networks to perform two tasks, including the prediction of enhancive/depressive DDIs and the prediction of directed DDIs. Based on DDI entries collected from DrugBank, the superiority of our model is demonstrated by the comparison with other state-of-the-art methods. Furthermore, the ablation study verifies that Balance and Status patterns help characterize directed pharmacological DDIs, and that the joint of two tasks provides better DDI representations than individual tasks. Last, we demonstrate the practical effectiveness of our model by a version-dependent test, where 88.47 and 81.38% DDI out of newly added entries provided by the latest release of DrugBank are validated in two predicting tasks respectively.

MOTIVATION: Biological experimental approaches to protein-protein interaction (PPI) site prediction are critical for understanding the mechanisms of biochemical processes but are time-consuming and laborious. With the development of Deep Learning (DL) techniques, the most popular Convolutional Neural Networks (CNN)-based methods have been proposed to address these problems. Although significant progress has been made, these methods still have limitations in encoding the characteristics of each amino acid in protein sequences. Current methods cannot efficiently explore the nature of Position Specific Scoring Matrix (PSSM), secondary structure and raw protein sequences by processing them all together. For PPI site prediction, how to effectively model the PPI context with attention to prediction remains an open problem. In addition, the long-distance dependencies of PPI features are important, which is very challenging for many CNN-based methods because the innate ability of CNN is difficult to outperform auto-regressive models like Transformers. RESULTS: To effectively mine the properties of PPI features, a novel hybrid neural network named HN-PPISP is proposed, which integrates a Multi-layer Perceptron Mixer (MLP-Mixer) module for local feature extraction and a two-stage multi-branch module for global feature capture. The model merits Transformer, TextCNN and Bi-LSTM as a powerful alternative for PPI site prediction. On the one hand, this is the first application of an advanced Transformer (i.e. MLP-Mixer) with a hybrid network for sequence-based PPI prediction. On the other hand, unlike existing methods that treat global features altogether, the proposed two-stage multi-branch hybrid module firstly assigns different attention scores to the input features and then encodes the feature through different branch modules. In the first stage, different improved attention modules are hybridized to extract features from the raw protein sequences, secondary structure and PSSM, respectively. In the second stage, a multi-branch network is designed to aggregate information from both branches in parallel. The two branches encode the features and extract dependencies through several operations such as TextCNN, Bi-LSTM and different activation functions. Experimental results on real-world public datasets show that our model consistently achieves state-of-the-art performance over seven remarkable baselines. AVAILABILITY: The source code of HN-PPISP model is available at https://github.com/ylxu05/HN-PPISP.

The advent of single-cell RNA-sequencing (scRNA-seq) provides an unprecedented opportunity to explore gene expression profiles at the single-cell level. However, gene expression values vary over time and under different conditions even within the same cell. There is an urgent need for more stable and reliable feature variables at the single-cell level to depict cell heterogeneity. Thus, we construct a new feature matrix called the delta rank matrix (DRM) from scRNA-seq data by integrating an a priori gene interaction network, which transforms the unreliable gene expression value into a stable gene interaction/edge value on a single-cell basis. This is the first time that a gene-level feature has been transformed into an interaction/edge-level for scRNA-seq data analysis based on relative expression orderings. Experiments on various scRNA-seq datasets have demonstrated that DRM performs better than the original gene expression matrix in cell clustering, cell identification and pseudo-trajectory reconstruction. More importantly, the DRM really achieves the fusion of gene expressions and gene interactions and provides a method of measuring gene interactions at the single-cell level. Thus, the DRM can be used to find changes in gene interactions among different cell types, which may open up a new way to analyze scRNA-seq data from an interaction perspective. In addition, DRM provides a new method to construct a cell-specific network for each single cell instead of a group of cells as in traditional network construction methods. DRM's exceptional performance is due to its extraction of rich gene-association information on biological systems and stable characterization of cells.

Lipidomics is of growing importance for clinical and biomedical research due to many associations between lipid metabolism and diseases. The discovery of these associations is facilitated by improved lipid identification and quantification. Sophisticated computational methods are advantageous for interpreting such large-scale data for understanding metabolic processes and their underlying (patho)mechanisms. To generate hypothesis about these mechanisms, the combination of metabolic networks and graph algorithms is a powerful option to pinpoint molecular disease drivers and their interactions. Here we present lipid network explorer (LINEX$^2$), a lipid network analysis framework that fuels biological interpretation of alterations in lipid compositions. By integrating lipid-metabolic reactions from public databases, we generate dataset-specific lipid interaction networks. To aid interpretation of these networks, we present an enrichment graph algorithm that infers changes in enzymatic activity in the context of their multispecificity from lipidomics data. Our inference method successfully recovered the MBOAT7 enzyme from knock-out data. Furthermore, we mechanistically interpret lipidomic alterations of adipocytes in obesity by leveraging network enrichment and lipid moieties. We address the general lack of lipidomics data mining options to elucidate potential disease mechanisms and make lipidomics more clinically relevant.

Genotype-by-environment interaction (GEI or GxE) plays an important role in understanding complex human traits. However, it is usually challenging to detect GEI signals efficiently and accurately while adjusting for population stratification and sample relatedness in large-scale genome-wide association studies (GWAS). Here we propose a fast and powerful linear mixed model-based approach, fastGWA-GE, to test for GEI effect and G + GxE joint effect. Our extensive simulations show that fastGWA-GE outperforms other existing GEI test methods by controlling genomic inflation better, providing larger power and running hundreds to thousands of times faster. We performed a fastGWA-GE analysis of ~7.27 million variants on 452 249 individuals of European ancestry for 13 quantitative traits and five environment variables in the UK Biobank GWAS data and identified 96 significant signals (72 variants across 57 loci) with GEI test P-values < 1 × 10-9, including 27 novel GEI associations, which highlights the effectiveness of fastGWA-GE in GEI signal discovery in large-scale GWAS.

Many enhancers exist as clusters in the genome and control cell identity and disease genes; however, the underlying mechanism remains largely unknown. Here, we introduce an algorithm, eNet, to build enhancer networks by integrating single-cell chromatin accessibility and gene expression profiles. The complexity of enhancer networks is assessed by two metrics: the number of enhancers and the frequency of predicted enhancer interactions (PEIs) based on chromatin co-accessibility. We apply eNet algorithm to a human blood dataset and find cell identity and disease genes tend to be regulated by complex enhancer networks. The network hub enhancers (enhancers with frequent PEIs) are the most functionally important. Compared with super-enhancers, enhancer networks show better performance in predicting cell identity and disease genes. eNet is robust and widely applicable in various human or mouse tissues datasets. Thus, we propose a model of enhancer networks containing three modes: Simple, Multiple and Complex, which are distinguished by their complexity in regulating gene expression. Taken together, our work provides an unsupervised approach to simultaneously identify key cell identity and disease genes and explore the underlying regulatory relationships among enhancers in single cells.

piRNA and PIWI proteins have been confirmed for disease diagnosis and treatment as novel biomarkers due to its abnormal expression in various cancers. However, the current research is not strong enough to further clarify the functions of piRNA in cancer and its underlying mechanism. Therefore, how to provide large-scale and serious piRNA candidates for biological research has grown up to be a pressing issue. In this study, a novel computational model based on the structural perturbation method is proposed to predict potential disease-associated piRNAs, called SPRDA. Notably, SPRDA belongs to positive-unlabeled learning, which is unaffected by negative examples in contrast to previous approaches. In the 5-fold cross-validation, SPRDA shows high performance on the benchmark dataset piRDisease, with an AUC of 0.9529. Furthermore, the predictive performance of SPRDA for 10 diseases shows the robustness of the proposed method. Overall, the proposed approach can provide unique insights into the pathogenesis of the disease and will advance the field of oncology diagnosis and treatment.

Single-cell ribonucleic acid (RNA) sequencing (scRNA-seq) has been a powerful technology for transcriptome analysis. However, the systematic validation of diverse computational tools used in scRNA-seq analysis remains challenging. Here, we propose a novel simulation tool, termed as Simulation of Cellular Heterogeneity (SimCH), for the flexible and comprehensive assessment of scRNA-seq computational methods. The Gaussian Copula framework is recruited to retain gene coexpression of experimental data shown to be associated with cellular heterogeneity. The synthetic count matrices generated by suitable SimCH modes closely match experimental data originating from either homogeneous or heterogeneous cell populations and either unique molecular identifier (UMI)-based or non-UMI-based techniques. We demonstrate how SimCH can benchmark several types of computational methods, including cell clustering, discovery of differentially expressed genes, trajectory inference, batch correction and imputation. Moreover, we show how SimCH can be used to conduct power evaluation of cell clustering methods. Given these merits, we believe that SimCH can accelerate single-cell research.

Genomic recombination is an important driving force for viral evolution, and recombination events have been reported for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) during the Coronavirus Disease 2019 pandemic, which significantly alter viral infectivity and transmissibility. However, it is difficult to identify viral recombination, especially for low-divergence viruses such as SARS-CoV-2, since it is hard to distinguish recombination from in situ mutation. Herein, we applied information theory to viral recombination analysis and developed VirusRecom, a program for efficiently screening recombination events on viral genome. In principle, we considered a recombination event as a transmission process of ``information'' and introduced weighted information content (WIC) to quantify the contribution of recombination to a certain region on viral genome; then, we identified the recombination regions by comparing WICs of different regions. In the benchmark using simulated data, VirusRecom showed a good balance between precision and recall compared to two competing tools, RDP5 and 3SEQ. In the detection of SARS-CoV-2 XE, XD and XF recombinants, VirusRecom providing more accurate positions of recombination regions than RDP5 and 3SEQ. In addition, we encapsulated the VirusRecom program into a command-line-interface software for convenient operation by users. In summary, we developed a novel approach based on information theory to identify viral recombination within highly similar sequences, providing a useful tool for monitoring viral evolution and epidemic control.

As microRNAs (miRNAs) are involved in many essential biological processes, their abnormal expressions can serve as biomarkers and prognostic indicators to prevent the development of complex diseases, thus providing accurate early detection and prognostic evaluation. Although a number of computational methods have been proposed to predict miRNA-disease associations (MDAs) for further experimental verification, their performance is limited primarily by the inadequacy of exploiting lower order patterns characterizing known MDAs to identify missing ones from MDA networks. Hence, in this work, we present a novel prediction model, namely HiSCMDA, by incorporating higher order network structures for improved performance of MDA prediction. To this end, HiSCMDA first integrates miRNA similarity network, disease similarity network and MDA network to preserve the advantages of all these networks. After that, it identifies overlapping functional modules from the integrated network by predefining several higher order connectivity patterns of interest. Last, a path-based scoring function is designed to infer potential MDAs based on network paths across related functional modules. HiSCMDA yields the best performance across all datasets and evaluation metrics in the cross-validation and independent validation experiments. Furthermore, in the case studies, 49 and 50 out of the top 50 miRNAs, respectively, predicted for colon neoplasms and lung neoplasms have been validated by well-established databases. Experimental results show that rich higher order organizational structures exposed in the MDA network gain new insight into the MDA prediction based on higher order connectivity patterns.

Cumulative studies have shown that many long non-coding RNAs (lncRNAs) are crucial in a number of diseases. Predicting potential lncRNA-disease associations (LDAs) can facilitate disease prevention, diagnosis and treatment. Therefore, it is vital to develop practical computational methods for LDA prediction. In this study, we propose a novel predictor named capsule network (CapsNet)-LDA for LDA prediction. CapsNet-LDA first uses a stacked autoencoder for acquiring the informative low-dimensional representations of the lncRNA-disease pairs under multiple views, then the attention mechanism is leveraged to implement an adaptive allocation of importance weights to them, and they are subsequently processed using a CapsNet-based architecture for predicting LDAs. Different from the conventional convolutional neural networks (CNNs) that have some restrictions with the usage of scalar neurons and pooling operations. the CapsNets use vector neurons instead of scalar neurons that have better robustness for the complex combination of features and they use dynamic routing processes for updating parameters. CapsNet-LDA is superior to other five state-of-the-art models on four benchmark datasets, four perturbed datasets and an independent test set in the comparison experiments, demonstrating that CapsNet-LDA has excellent performance and robustness against perturbation, as well as good generalization ability. The ablation studies verify the effectiveness of some modules of CapsNet-LDA. Moreover, the ability of multi-view data to improve performance is proven. Case studies further indicate that CapsNet-LDA can accurately predict novel LDAs for specific diseases.

Dimension reduction (DR) plays an important role in single-cell RNA sequencing (scRNA-seq), such as data interpretation, visualization and other downstream analysis. A desired DR method should be applicable to various application scenarios, including identifying cell types, preserving the inherent structure of data and handling with batch effects. However, most of the existing DR methods fail to accommodate these requirements simultaneously, especially removing batch effects. In this paper, we develop a novel structure-preserved dimension reduction (SPDR) method using intra- and inter-batch triplets sampling. The constructed triplets jointly consider each anchor's mutual nearest neighbors from inter-batch, k-nearest neighbors from intra-batch and randomly selected cells from the whole data, which capture higher order structure information and meanwhile account for batch information of the data. Then we minimize a robust loss function for the chosen triplets to obtain a structure-preserved and batch-corrected low-dimensional representation. Comprehensive evaluations show that SPDR outperforms other competing DR methods, such as INSCT, IVIS, Trimap, Scanorama, scVI and UMAP, in removing batch effects, preserving biological variation, facilitating visualization and improving clustering accuracy. Besides, the two-dimensional (2D) embedding of SPDR presents a clear and authentic expression pattern, and can guide researchers to determine how many cell types should be identified. Furthermore, SPDR is robust to complex data characteristics (such as down-sampling, duplicates and outliers) and varying hyperparameter settings. We believe that SPDR will be a valuable tool for characterizing complex cellular heterogeneity.

Single-cell RNA-sequencing technology (scRNA-seq) brings research to single-cell resolution. However, a major drawback of scRNA-seq is large sparsity, i.e. expressed genes with no reads due to technical noise or limited sequence depth during the scRNA-seq protocol. This phenomenon is also called 'dropout' events, which likely affect downstream analyses such as differential expression analysis, the clustering and visualization of cell subpopulations, cellular trajectory inference, etc. Therefore, there is a need to develop a method to identify and impute these dropout events. We propose Bubble, which first identifies dropout events from all zeros based on expression rate and coefficient of variation of genes within cell subpopulation, and then leverages an autoencoder constrained by bulk RNA-seq data to only impute those values. Unlike other deep learning-based imputation methods, Bubble fuses the matched bulk RNA-seq data as a constraint to reduce the introduction of false positive signals. Using simulated and several real scRNA-seq datasets, we demonstrate that Bubble enhances the recovery of missing values, gene-to-gene and cell-to-cell correlations, and reduces the introduction of false positive signals. Regarding some crucial downstream analyses of scRNA-seq data, Bubble facilitates the identification of differentially expressed genes, improves the performance of clustering and visualization, and aids the construction of cellular trajectory. More importantly, Bubble provides fast and scalable imputation with minimal memory usage.

The development of targeted drugs allows precision medicine in cancer treatment and optimal targeted therapies. Accurate identification of cancer druggable genes helps strengthen the understanding of targeted cancer therapy and promotes precise cancer treatment. However, rare cancer-druggable genes have been found due to the multi-omics data's diversity and complexity. This study proposes deep forest for cancer druggable genes discovery (DF-CAGE), a novel machine learning-based method for cancer-druggable gene discovery. DF-CAGE integrated the somatic mutations, copy number variants, DNA methylation and RNA-Seq data across ˜10 000 TCGA profiles to identify the landscape of the cancer-druggable genes. We found that DF-CAGE discovers the commonalities of currently known cancer-druggable genes from the perspective of multi-omics data and achieved excellent performance on OncoKB, Target and Drugbank data sets. Among the ˜20 000 protein-coding genes, DF-CAGE pinpointed 465 potential cancer-druggable genes. We found that the candidate cancer druggable genes (CDG) are clinically meaningful and divided the CDG into known, reliable and potential gene sets. Finally, we analyzed the omics data's contribution to identifying druggable genes. We found that DF-CAGE reports druggable genes mainly based on the copy number variations (CNVs) data, the gene rearrangements and the mutation rates in the population. These findings may enlighten the future study and development of new drugs.

BACKGROUND: Global or untargeted metabolomics is widely used to comprehensively investigate metabolic profiles under various pathophysiological conditions such as inflammations, infections, responses to exposures or interactions with microbial communities. However, biological interpretation of global metabolomics data remains a daunting task. Recent years have seen growing applications of pathway enrichment analysis based on putative annotations of liquid chromatography coupled with mass spectrometry (LC-MS) peaks for functional interpretation of LC-MS-based global metabolomics data. However, due to intricate peak-metabolite and metabolite-pathway relationships, considerable variations are observed among results obtained using different approaches. There is an urgent need to benchmark these approaches to inform the best practices. RESULTS: We have conducted a benchmark study of common peak annotation approaches and pathway enrichment methods in current metabolomics studies. Representative approaches, including three peak annotation methods and four enrichment methods, were selected and benchmarked under different scenarios. Based on the results, we have provided a set of recommendations regarding peak annotation, ranking metrics and feature selection. The overall better performance was obtained for the mummichog approach. We have observed that a ~30% annotation rate is sufficient to achieve high recall (~90% based on mummichog), and using semi-annotated data improves functional interpretation. Based on the current platforms and enrichment methods, we further propose an identifiability index to indicate the possibility of a pathway being reliably identified. Finally, we evaluated all methods using 11 COVID-19 and 8 inflammatory bowel diseases (IBD) global metabolomics datasets.

Biomedical multi-modality data (also named multi-omics data) refer to data that span different types and derive from multiple sources in clinical practices (e.g. gene sequences, proteomics and histopathological images), which can provide comprehensive perspectives for cancers and generally improve the performance of survival models. However, the performance improvement of multi-modality survival models may be hindered by two key issues as follows: (1) how to learn and fuse modality-sharable and modality-individual representations from multi-modality data; (2) how to explore the potential risk-aware characteristics in each risk subgroup, which is beneficial to risk stratification and prognosis evaluation. Additionally, learning-based survival models generally refer to numerous hyper-parameters, which requires time-consuming parameter setting and might result in a suboptimal solution. In this paper, we propose an adaptive risk-aware sharable and individual subspace learning method for cancer survival analysis. The proposed method jointly learns sharable and individual subspaces from multi-modality data, whereas two auxiliary terms (i.e. intra-modality complementarity and inter-modality incoherence) are developed to preserve the complementary and distinctive properties of each modality. Moreover, it equips with a grouping co-expression constraint for obtaining risk-aware representation and preserving local consistency. Furthermore, an adaptive-weighted strategy is employed to efficiently estimate crucial parameters during the training stage. Experimental results on three public datasets demonstrate the superiority of our proposed model.

One of key features of intrinsically disordered regions (IDRs) is facilitation of protein-protein and protein-nucleic acids interactions. These disordered binding regions include molecular recognition features (MoRFs), short linear motifs (SLiMs) and longer binding domains. Vast majority of current predictors of disordered binding regions target MoRFs, with a handful of methods that predict SLiMs and disordered protein-binding domains. A new and broader class of disordered binding regions, linear interacting peptides (LIPs), was introduced recently and applied in the MobiDB resource. LIPs are segments in protein sequences that undergo disorder-to-order transition upon binding to a protein or a nucleic acid, and they cover MoRFs, SLiMs and disordered protein-binding domains. Although current predictors of MoRFs and disordered protein-binding regions could be used to identify some LIPs, there are no dedicated sequence-based predictors of LIPs. To this end, we introduce CLIP, a new predictor of LIPs that utilizes robust logistic regression model to combine three complementary types of inputs: co-evolutionary information derived from multiple sequence alignments, physicochemical profiles and disorder predictions. Ablation analysis suggests that the co-evolutionary information is particularly useful for this prediction and that combining the three inputs provides substantial improvements when compared to using these inputs individually. Comparative empirical assessments using low-similarity test datasets reveal that CLIP secures area under receiver operating characteristic curve (AUC) of 0.8 and substantially improves over the results produced by the closest current tools that predict MoRFs and disordered protein-binding regions. The webserver of CLIP is freely available at http://biomine.cs.vcu.edu/servers/CLIP/ and the standalone code can be downloaded from http://yanglab.qd.sdu.edu.cn/download/CLIP/.

Erratum in Brief Bioinform. 2023 Jan 09;:

There are a number of antigens that induce autoimmune response against β-cells, leading to type 1 diabetes mellitus (T1DM). Recently, several antigen-specific immunotherapies have been developed to treat T1DM. Thus, identification of T1DM associated peptides with antigenic regions or epitopes is important for peptide based-therapeutics (e.g. immunotherapeutic). In this study, for the first time, an attempt has been made to develop a method for predicting, designing, and scanning of T1DM associated peptides with high precision. We analysed 815 T1DM associated peptides and observed that these peptides are not associated with a specific class of HLA alleles. Thus, HLA binder prediction methods are not suitable for predicting T1DM associated peptides. First, we developed a similarity/alignment based method using Basic Local Alignment Search Tool and achieved a high probability of correct hits with poor coverage. Second, we developed an alignment-free method using machine learning techniques and got a maximum AUROC of 0.89 using dipeptide composition. Finally, we developed a hybrid method that combines the strength of both alignment free and alignment-based methods and achieves maximum area under the receiver operating characteristic of 0.95 with Matthew's correlation coefficient of 0.81 on an independent dataset. We developed a web server 'DMPPred' and stand-alone server for predicting, designing and scanning T1DM associated peptides (https://webs.iiitd.edu.in/raghava/dmppred/).

LTR-retrotransposons are the most abundant repeat sequences in plant genomes and play an important role in evolution and biodiversity. Their characterization is of great importance to understand their dynamics. However, the identification and classification of these elements remains a challenge today. Moreover, current software can be relatively slow (from hours to days), sometimes involve a lot of manual work and do not reach satisfactory levels in terms of precision and sensitivity. Here we present Inpactor2, an accurate and fast application that creates LTR-retrotransposon reference libraries in a very short time. Inpactor2 takes an assembled genome as input and follows a hybrid approach (deep learning and structure-based) to detect elements, filter partial sequences and finally classify intact sequences into superfamilies and, as very few tools do, into lineages. This tool takes advantage of multi-core and GPU architectures to decrease execution times. Using the rice genome, Inpactor2 showed a run time of 5 minutes (faster than other tools) and has the best accuracy and F1-Score of the tools tested here, also having the second best accuracy and specificity only surpassed by EDTA, but achieving 28% higher sensitivity. For large genomes, Inpactor2 is up to seven times faster than other available bioinformatics tools.

Chromosomal translocations in cancer genomes, key players in many types of cancers, generate chimeric proteins that drive oncogenesis. Genomes with chromosomal rearrangements can also produce fusion circular RNAs (f-circRNAs) by backsplicing of chimeric transcripts, as first shown in leukemias with PML::RARα and KMT2A::MLLT3 translocations and later in solid cancers. F-circRNAs contribute to the oncogenic processes and reinforce the oncogenic activity of chimeric proteins. In leukemia with KMT2A::AFF1 (MLL::AF4) fusions, we previously reported specific alterations of circRNA expression, but nothing was known about f-circRNAs. Due to the presence of two chimeric sequences, fusion and backsplice junctions, the identification of f-circRNAs with available tools is challenging, possibly resulting in the underestimation of this RNA species, especially when the breakpoint is not known. We developed CircFusion, a new software tool to detect linear fusion transcripts and f-circRNAs from RNA-seq data, both in samples for which the breakpoints are known and when the information about the joined exons is missing. CircFusion can detect linear and circular chimeric transcripts deriving from the main and reciprocal translocations also in the presence of multiple breakpoints, which are common in malignant cells. Benchmarking tests on simulated and real datasets of cancer samples with previously experimentally determined f-circRNAs showed that CircFusion provides reliable predictions and outperforms available methods for f-circRNA detection. We discovered and validated novel f-circRNAs in acute leukemia harboring KMT2A::AFF1 rearrangements, leading the way to future functional studies aimed to unveil their role in this malignancy.

Protein language models, trained on millions of biologically observed sequences, generate feature-rich numerical representations of protein sequences. These representations, called sequence embeddings, can infer structure-functional properties, despite protein language models being trained on primary sequence alone. While sequence embeddings have been applied toward tasks such as structure and function prediction, applications toward alignment-free sequence classification have been hindered by the lack of studies to derive, quantify and evaluate relationships between protein sequence embeddings. Here, we develop workflows and visualization methods for the classification of protein families using sequence embedding derived from protein language models. A benchmark of manifold visualization methods reveals that Neighbor Joining (NJ) embedding trees are highly effective in capturing global structure while achieving similar performance in capturing local structure compared with popular dimensionality reduction techniques such as t-SNE and UMAP. The statistical significance of hierarchical clusters on a tree is evaluated by resampling embeddings using a variational autoencoder (VAE). We demonstrate the application of our methods in the classification of two well-studied enzyme superfamilies, phosphatases and protein kinases. Our embedding-based classifications remain consistent with and extend upon previously published sequence alignment-based classifications. We also propose a new hierarchical classification for the S-Adenosyl-L-Methionine (SAM) enzyme superfamily which has been difficult to classify using traditional alignment-based approaches. Beyond applications in sequence classification, our results further suggest NJ trees are a promising general method for visualizing high-dimensional data sets.

Cancer initiation and progression are likely caused by the dysregulation of biological pathways. Gene set analysis (GSA) could improve the signal-to-noise ratio and identify potential biological insights on the gene set level. However, platforms exploring cancer multi-omics data using GSA methods are lacking. In this study, we upgraded our GSCALite to GSCA (gene set cancer analysis, http://bioinfo.life.hust.edu.cn/GSCA) for cancer GSA at genomic, pharmacogenomic and immunogenomic levels. In this improved GSCA, we integrated expression, mutation, drug sensitivity and clinical data from four public data sources for 33 cancer types. We introduced useful features to GSCA, including associations between immune infiltration with gene expression and genomic variations, and associations between gene set expression/mutation and clinical outcomes. GSCA has four main functional modules for cancer GSA to explore, analyze and visualize expression, genomic variations, tumor immune infiltration, drug sensitivity and their associations with clinical outcomes. We used case studies of three gene sets: (i) seven cell cycle genes, (ii) tumor suppressor genes of PI3K pathway and (iii) oncogenes of PI3K pathway to prove the advantage of GSCA over single gene analysis. We found novel associations of gene set expression and mutation with clinical outcomes in different cancer types on gene set level, while on single gene analysis level, they are not significant associations. In conclusion, GSCA is a user-friendly web server and a useful resource for conducting hypothesis tests by using GSA methods at genomic, pharmacogenomic and immunogenomic levels.

The rapid progress of machine learning (ML) in predicting molecular properties enables high-precision predictions being routinely achieved. However, many ML models, such as conventional molecular graph, cannot differentiate stereoisomers of certain types, particularly conformational and chiral ones that share the same bonding connectivity but differ in spatial arrangement. Here, we designed a hybrid molecular graph network, Chemical Feature Fusion Network (CFFN), to address the issue by integrating planar and stereo information of molecules in an interweaved fashion. The three-dimensional (3D, i.e., stereo) modality guarantees precision and completeness by providing unabridged information, while the two-dimensional (2D, i.e., planar) modality brings in chemical intuitions as prior knowledge for guidance. The zipper-like arrangement of 2D and 3D information processing promotes cooperativity between them, and their synergy is the key to our model's success. Experiments on various molecules or conformational datasets including a special newly created chiral molecule dataset comprised of various configurations and conformations demonstrate the superior performance of CFFN. The advantage of CFFN is even more significant in datasets made of small samples. Ablation experiments confirm that fusing 2D and 3D molecular graphs as unambiguous molecular descriptors can not only effectively distinguish molecules and their conformations, but also achieve more accurate and robust prediction of quantum chemical properties.

Pathway genes functionally participate in the same biological process. They typically act cooperatively, and none is considered dispensable. The dominant paradigm in drug discovery is the one-to-one strategy, which aims to find the most sensitive drug to act on an individual target. However, many complex diseases, such as cancer, are caused by dysfunction among multiple-gene pathways, not just one. Therefore, identifying pathway genes that are responsive to synthetic compounds in a global physiological environment may be more effective in drug discovery. The high redundancy of crosstalk between biological pathways, though, hints that the covariance matrix, which only connects genes with strong marginal correlations, may miss higher-level interactions, such as group interactions. We herein report the development of DPADM-a Drug-Pathway association Detection Model that infers pathways responsive to specific drugs. This model elucidates higher-level gene-gene interactions by evaluating the conditional dependencies between genes under different drug treatments. The advantage of the proposed method is demonstrated using simulation studies by comparing with another two methods. We applied this model to the Connectivity Map data set (CMap), and demonstrated that DPADM is able to identify many drug-pathway associations, such as mitoxantrone (MTX)- PI3K/AKT association, which targets the topological conditions of DNA transcription. Surprisingly, apart from identifying pathways corresponding to specific drugs, our methodology also revealed new drug-related pathways with functions similarly to those of seed genes.

Trans-splicing of a spliced leader (SL) to the 5' ends of mRNAs is used to produce mature mRNAs in several phyla of great importance to human health and the marine ecosystem. One of the consequences of the addition of SL sequences is the change or disruption of the open reading frames (ORFs) in the recipient transcripts. Given that most SL sequences have one or more of the trinucleotide NUG, including AUG in flatworms, trans-splicing of SL sequences can potentially supply a start codon to create new ORFs, which we refer to as slORFs, in the recipient mRNAs. Due to the lack of a tool to precisely detect them, slORFs were usually neglected in previous studies. In this work, we present the tool slORFfinder, which automatically links the SL sequences to the recipient mRNAs at the trans-splicing sites identified from SL-containing reads of RNA-Seq and predicts slORFs according to the distribution of ribosome-protected footprints (RPFs) on the trans-spliced transcripts. By applying this tool to the analyses of nematodes, ascidians and euglena, whose RPFs are publicly available, we find wide existence of slORFs in these taxa. Furthermore, we find that slORFs are generally translated at higher levels than the annotated ORFs in the genomes, suggesting they might have important functions. Overall, this study provides a tool, slORFfinder (https://github.com/songbo446/slORFfinder), to identify slORFs, which can enhance our understanding of ORFs in taxa with SL machinery.

Drug response prediction is an important problem in personalized cancer therapy. Among various newly developed models, significant improvement in prediction performance has been reported using deep learning methods. However, systematic comparisons of deep learning methods, especially of the transferability from preclinical models to clinical cohorts, are currently lacking. To provide a more rigorous assessment, the performance of six representative deep learning methods for drug response prediction using nine evaluation metrics, including the overall prediction accuracy, predictability of each drug, potential associated factors and transferability to clinical cohorts, in multiple application scenarios was benchmarked. Most methods show promising prediction within cell line datasets, and TGSA, with its lower time cost and better performance, is recommended. Although the performance metrics decrease when applying models trained on cell lines to patients, a certain amount of power to distinguish clinical response on some drugs can be maintained using CRDNN and TGSA. With these assessments, we provide a guidance for researchers to choose appropriate methods, as well as insights into future directions for the development of more effective methods in clinical scenarios.

Computational cell type deconvolution on bulk transcriptomics data can reveal cell type proportion heterogeneity across samples. One critical factor for accurate deconvolution is the reference signature matrix for different cell types. Compared with inferring reference signature matrices from cell lines, rapidly accumulating single-cell RNA-sequencing (scRNA-seq) data provide a richer and less biased resource. However, deriving cell type signature from scRNA-seq data is challenging due to high biological and technical noises. In this article, we introduce a novel Bayesian framework, tranSig, to improve signature matrix inference from scRNA-seq by leveraging shared cell type-specific expression patterns across different tissues and studies. Our simulations show that tranSig is robust to the number of signature genes and tissues specified in the model. Applications of tranSig to bulk RNA sequencing data from peripheral blood, bronchoalveolar lavage and aorta demonstrate its accuracy and power to characterize biological heterogeneity across groups. In summary, tranSig offers an accurate and robust approach to defining gene expression signatures of different cell types, facilitating improved in silico cell type deconvolutions.

Immune checkpoint inhibitors (ICI) show high efficiency in a small fraction of advanced gastric cancer (GC). However, personalized immune subtypes have not been developed for the prediction of ICI efficiency in GC. Herein, we identified Pan-Immune Activation Module (PIAM), a curated gene expression profile (GEP) representing the co-infiltration of multiple immune cell types in tumor microenvironment of GC, which was associated with high expression of immunosuppressive molecules such as PD-1 and CTLA-4. We also identified Pan-Immune Dysfunction Genes (PIDG), a conservative PIAM-derivated GEP indicating the dysfunction of immune cell cooperation, which was associated with upregulation of metastatic programs (extracellular matrix receptor interaction, TGF-β signaling, epithelial-mesenchymal transition and calcium signaling) but downregulation of proliferative signalings (MYC targets, E2F targets, mTORC1 signaling, and DNA replication and repair). Moreover, we developed 'GSClassifier', an ensemble toolkit based on top scoring pairs and extreme gradient boosting, for population-based modeling and personalized identification of GEP subtypes. With PIAM and PIDG, we developed four Pan-immune Activation and Dysfunction (PAD) subtypes and a GSClassifier model 'PAD for individual' with high accuracy in predicting response to pembrolizumab (anti-PD-1) in advance GC (AUC = 0.833). Intriguingly, PAD-II (PIAMhighPIDGlow) displayed the highest objective response rate (60.0%) compared with other subtypes (PAD-I, PIAMhighPIDGhigh, 0%; PAD-III, PIAMlowPIDGhigh, 0%; PAD-IV, PIAMlowPIDGlow, 17.6%; P = 0.003), which was further validated in the metastatic urothelial cancer cohort treated with atezolizumab (anti-PD-L1) (P = 0.018). In all, we provided 'GSClassifier' as a refined computational framework for GEP-based stratification and PAD subtypes as a promising strategy for exploring ICI responders in GC. Metastatic pathways could be potential targets for GC patients with high immune infiltration but resistance to ICI therapy.

The time since deposition (TSD) of a bloodstain, i.e., the time of a bloodstain formation is an essential piece of biological evidence in crime scene investigation. The practical usage of some existing microscopic methods (e.g., spectroscopy or RNA analysis technology) is limited, as their performance strongly relies on high-end instrumentation and/or rigorous laboratory conditions. This paper presents a practically applicable deep learning-based method (i.e., BloodNet) for efficient, accurate, and costless TSD inference from a macroscopic view, i.e., by using easily accessible bloodstain photos. To this end, we established a benchmark database containing around 50,000 photos of bloodstains with varying TSDs. Capitalizing on such a large-scale database, BloodNet adopted attention mechanisms to learn from relatively high-resolution input images the localized fine-grained feature representations that were highly discriminative between different TSD periods. Also, the visual analysis of the learned deep networks based on the Smooth Grad-CAM tool demonstrated that our BloodNet can stably capture the unique local patterns of bloodstains with specific TSDs, suggesting the efficacy of the utilized attention mechanism in learning fine-grained representations for TSD inference. As a paired study for BloodNet, we further conducted a microscopic analysis using Raman spectroscopic data and a machine learning method based on Bayesian optimization. Although the experimental results show that such a new microscopic-level approach outperformed the state-of-the-art by a large margin, its inference accuracy is significantly lower than BloodNet, which further justifies the efficacy of deep learning techniques in the challenging task of bloodstain TSD inference. Our code is publically accessible via https://github.com/shenxiaochenn/BloodNet. Our datasets and pre-trained models can be freely accessed via https://figshare.com/articles/dataset/21291825.

Accurate prediction of deoxyribonucleic acid (DNA) modifications is essential to explore and discern the process of cell differentiation, gene expression and epigenetic regulation. Several computational approaches have been proposed for particular type-specific DNA modification prediction. Two recent generalized computational predictors are capable of detecting three different types of DNA modifications; however, type-specific and generalized modifications predictors produce limited performance across multiple species mainly due to the use of ineffective sequence encoding methods. The paper in hand presents a generalized computational approach "DNA-MP" that is competent to more precisely predict three different DNA modifications across multiple species. Proposed DNA-MP approach makes use of a powerful encoding method "position specific nucleotides occurrence based 117 on modification and non-modification class densities normalized difference" (POCD-ND) to generate the statistical representations of DNA sequences and a deep forest classifier for modifications prediction. POCD-ND encoder generates statistical representations by extracting position specific distributional information of nucleotides in the DNA sequences. We perform a comprehensive intrinsic and extrinsic evaluation of the proposed encoder and compare its performance with 32 most widely used encoding methods on $17$ benchmark DNA modifications prediction datasets of $12$ different species using $10$ different machine learning classifiers. Overall, with all classifiers, the proposed POCD-ND encoder outperforms existing $32$ different encoders. Furthermore, combinedly over 5-fold cross validation benchmark datasets and independent test sets, proposed DNA-MP predictor outperforms state-of-the-art type-specific and generalized modifications predictors by an average accuracy of 7% across 4mc datasets, 1.35% across 5hmc datasets and 10% for 6ma datasets. To facilitate the scientific community, the DNA-MP web application is available at https://sds_genetic_analysis.opendfki.de/DNA_Modifications/.

The subcellular localization of long non-coding RNAs (lncRNAs) is crucial for understanding lncRNA functions. Most of existing lncRNA subcellular localization prediction methods use k-mer frequency features to encode lncRNA sequences. However, k-mer frequency features lose sequence order information and fail to capture sequence patterns and motifs of different lengths. In this paper, we proposed GraphLncLoc, a graph convolutional network-based deep learning model, for predicting lncRNA subcellular localization. Unlike previous studies encoding lncRNA sequences by using k-mer frequency features, GraphLncLoc transforms lncRNA sequences into de Bruijn graphs, which transforms the sequence classification problem into a graph classification problem. To extract the high-level features from the de Bruijn graph, GraphLncLoc employs graph convolutional networks to learn latent representations. Then, the high-level feature vectors derived from de Bruijn graph are fed into a fully connected layer to perform the prediction task. Extensive experiments show that GraphLncLoc achieves better performance than traditional machine learning models and existing predictors. In addition, our analyses show that transforming sequences into graphs has more distinguishable features and is more robust than k-mer frequency features. The case study shows that GraphLncLoc can uncover important motifs for nucleus subcellular localization. GraphLncLoc web server is available at http://csuligroup.com:8000/GraphLncLoc/.

Deoxyribonucleic acid (DNA) is an attractive medium for long-term digital data storage due to its extremely high storage density, low maintenance cost and longevity. However, during the process of synthesis, amplification and sequencing of DNA sequences with homopolymers of large run-length, three different types of errors, namely, insertion, deletion and substitution errors frequently occur. Meanwhile, DNA sequences with large imbalances between GC and AT content exhibit high dropout rates and are prone to errors. These limitations severely hinder the widespread use of DNA-based data storage. In order to reduce and correct these errors in DNA storage, this paper proposes a novel coding schema called DNA-LC, which converts binary sequences into DNA base sequences that satisfy both the GC balance and run-length constraints. Furthermore, our coding mode is able to detect and correct multiple errors with a higher error correction capability than the other methods targeting single error correction within a single strand. The decoding algorithm has been implemented in practice. Simulation results indicate that our proposed coding scheme can offer outstanding error protection to DNA sequences. The source code is freely accessible at https://github.com/XiayangLi2301/DNA.

Viral infection involves a large number of protein-protein interactions (PPIs) between the virus and the host, and the identification of these PPIs plays an important role in revealing viral infection and pathogenesis. Existing computational models focus on predicting whether human proteins and viral proteins interact, and rarely take into account the types of diseases associated with these interactions. Although there are computational models based on a matrix and tensor decomposition for predicting multi-type biological interaction relationships, these methods cannot effectively model high-order nonlinear relationships of biological entities and are not suitable for integrating multiple features. To this end, we propose a novel computational framework, LTDSSL, to determine human-virus PPIs under different disease types. LTDSSL utilizes logistic functions to model nonlinear associations, sets importance levels to emphasize the importance of observed interactions and utilizes sparse subspace learning of multiple features to improve model performance. Experimental results show that LTDSSL has better predictive performance for both new disease types and new triples than the state-of-the-art methods. In addition, the case study further demonstrates that LTDSSL can effectively predict human-viral PPIs under various disease types.

High-throughput profiling of microbial functional traits involved in various biogeochemical cycling pathways using shotgun metagenomic sequencing has been routinely applied in microbial ecology and environmental science. Multiple bioinformatics data processing approaches are available, including assembly-based (single-sample assembly and multi-sample assembly) and read-based (merged reads and raw data). However, it remains not clear how these different approaches may differ in data analyses and affect result interpretation. In this study, using two typical shotgun metagenome datasets recovered from geographically distant coastal sediments, the performance of different data processing approaches was comparatively investigated from both technical and biological/ecological perspectives. Microbially mediated biogeochemical cycling pathways, including nitrogen cycling, sulfur cycling and B12 biosynthesis, were analyzed. As a result, multi-sample assembly provided the most amount of usable information for targeted functional traits, at a high cost of computational resources and running time. Single-sample assembly and read-based analysis were comparable in obtaining usable information, but the former was much more time- and resource-consuming. Critically, different approaches introduced much stronger variations in microbial profiles than biological differences. However, community-level differences between the two sampling sites could be consistently observed despite the approaches being used. In choosing an appropriate approach, researchers shall balance the trade-offs between multiple factors, including the scientific question, the amount of usable information, computational resources and time cost. This study is expected to provide valuable technical insights and guidelines for the various approaches used for metagenomic data analysis.

The recently reported machine learning- or deep learning-based scoring functions (SFs) have shown exciting performance in predicting protein-ligand binding affinities with fruitful application prospects. However, the differentiation between highly similar ligand conformations, including the native binding pose (the global energy minimum state), remains challenging that could greatly enhance the docking. In this work, we propose a fully differentiable, end-to-end framework for ligand pose optimization based on a hybrid SF called DeepRMSD+Vina combined with a multi-layer perceptron (DeepRMSD) and the traditional AutoDock Vina SF. The DeepRMSD+Vina, which combines (1) the root mean square deviation (RMSD) of the docking pose with respect to the native pose and (2) the AutoDock Vina score, is fully differentiable; thus is capable of optimizing the ligand binding pose to the energy-lowest conformation. Evaluated by the CASF-2016 docking power dataset, the DeepRMSD+Vina reaches a success rate of 94.4%, which outperforms most reported SFs to date. We evaluated the ligand conformation optimization framework in practical molecular docking scenarios (redocking and cross-docking tasks), revealing the high potentialities of this framework in drug design and discovery. Structural analysis shows that this framework has the ability to identify key physical interactions in protein-ligand binding, such as hydrogen-bonding. Our work provides a paradigm for optimizing ligand conformations based on deep learning algorithms. The DeepRMSD+Vina model and the optimization framework are available at GitHub repository https://github.com/zchwang/DeepRMSD-Vina_Optimization.

Recent studies have revealed that long noncoding RNAs (lncRNAs) are closely linked to several human diseases, providing new opportunities for their use in detection and therapy. Many graph propagation and similarity fusion approaches can be used for predicting potential lncRNA-disease associations. However, existing similarity fusion approaches suffer from noise and self-similarity loss in the fusion process. To address these problems, a new prediction approach, termed SSMF-BLNP, based on organically combining selective similarity matrix fusion (SSMF) and bidirectional linear neighborhood label propagation (BLNP), is proposed in this paper to predict lncRNA-disease associations. In SSMF, self-similarity networks of lncRNAs and diseases are obtained by selective preprocessing and nonlinear iterative fusion. The fusion process assigns weights to each initial similarity network and introduces a unit matrix that can reduce noise and compensate for the loss of self-similarity. In BLNP, the initial lncRNA-disease associations are employed in both lncRNA and disease directions as label information for linear neighborhood label propagation. The propagation was then performed on the self-similarity network obtained from SSMF to derive the scoring matrix for predicting the relationships between lncRNAs and diseases. Experimental results showed that SSMF-BLNP performed better than seven other state of-the-art approaches. Furthermore, a case study demonstrated up to 100% and 80% accuracy in 10 lncRNAs associated with hepatocellular carcinoma and 10 lncRNAs associated with renal cell carcinoma, respectively. The source code and datasets used in this paper are available at: https://github.com/RuiBingo/SSMF-BLNP.

MOTIVATION: Molecular property prediction is a significant requirement in AI-driven drug design and discovery, aiming to predict the molecular property information (e.g. toxicity) based on the mined biomolecular knowledge. Although graph neural networks have been proven powerful in predicting molecular property, unbalanced labeled data and poor generalization capability for new-synthesized molecules are always key issues that hinder further improvement of molecular encoding performance. RESULTS: We propose a novel self-supervised representation learning scheme based on a Cascaded Attention Network and Graph Contrastive Learning (CasANGCL). We design a new graph network variant, designated as cascaded attention network, to encode local-global molecular representations. We construct a two-stage contrast predictor framework to tackle the label imbalance problem of training molecular samples, which is an integrated end-to-end learning scheme. Moreover, we utilize the information-flow scheme for training our network, which explicitly captures the edge information in the node/graph representations and obtains more fine-grained knowledge. Our model achieves an 81.9% ROC-AUC average performance on 661 tasks from seven challenging benchmarks, showing better portability and generalizations. Further visualization studies indicate our model's better representation capacity and provide interpretability.

Unlike the typical single circular structure of most animal mitochondrial genomes (mitogenome), the drastic structural variation of plant mitogenomes is a result of a mixture of molecules of various sizes and structures. Obtaining the full panoramic plant mitogenome is still considered a roadblock in evolutionary biology. In this study, we developed a graph-based sequence assembly toolkit (GSAT) to construct the pan-structural landscape of plant mitogenome with high-quality mitochondrial master graphs (MMGs) for model species including rice (Oryza sativa) and thale cress (Arabidopsis thaliana). The rice and thale cress MMGs have total lengths of 346 562 and 358 041 bp, including 9 and 6 contigs and 12 and 8 links, respectively, and could be further divided into 6 and 3 minimum master circles and 4 and 2 minimum secondary circles separately. The nuclear mitochondrial DNA segments (NUMTs) in thale cress strongly affected the frequency evaluation of the homologous structures in the mitogenome, while the effects of NUMTs in rice were relatively weak. The mitochondrial plastid DNA segments (MTPTs) in both species had no effects on the assessment of the MMGs. All potential recombinant structures were evaluated, and the findings revealed that all, except for nuclear-homologous structures, MMG structures are present at a much higher frequency than non-MMG structures are. Investigations of potential circular and linear molecules further supported multiple dominant structures in the mitogenomes and could be completely summarized in the MMG. Our study provided an efficient and accurate model for assembling and applying graph-based plant mitogenomes to assess their pan-structural variations.

Ultraliser is a neuroscience-specific software framework capable of creating accurate and biologically realistic 3D models of complex neuroscientific structures at intracellular (e.g. mitochondria and endoplasmic reticula), cellular (e.g. neurons and glia) and even multicellular scales of resolution (e.g. cerebral vasculature and minicolumns). Resulting models are exported as triangulated surface meshes and annotated volumes for multiple applications in in silico neuroscience, allowing scalable supercomputer simulations that can unravel intricate cellular structure-function relationships. Ultraliser implements a high-performance and unconditionally robust voxelization engine adapted to create optimized watertight surface meshes and annotated voxel grids from arbitrary non-watertight triangular soups, digitized morphological skeletons or binary volumetric masks. The framework represents a major leap forward in simulation-based neuroscience, making it possible to employ high-resolution 3D structural models for quantification of surface areas and volumes, which are of the utmost importance for cellular and system simulations. The power of Ultraliser is demonstrated with several use cases in which hundreds of models are created for potential application in diverse types of simulations. Ultraliser is publicly released under the GNU GPL3 license on GitHub (BlueBrain/Ultraliser). SIGNIFICANCE: There is crystal clear evidence on the impact of cell shape on its signaling mechanisms. Structural models can therefore be insightful to realize the function; the more realistic the structure can be, the further we get insights into the function. Creating realistic structural models from existing ones is challenging, particularly when needed for detailed subcellular simulations. We present Ultraliser, a neuroscience-dedicated framework capable of building these structural models with realistic and detailed cellular geometries that can be used for simulations.

At present, the study on the pathogenesis of Alzheimer's disease (AD) by multimodal data fusion analysis has been attracted wide attention. It often has the problems of small sample size and high dimension with the multimodal medical data. In view of the characteristics of multimodal medical data, the existing genetic evolution random neural network cluster (GERNNC) model combine genetic evolution algorithm and neural network for the classification of AD patients and the extraction of pathogenic factors. However, the model does not take into account the non-linear relationship between brain regions and genes and the problem that the genetic evolution algorithm can fall into local optimal solutions, which leads to the overall performance of the model is not satisfactory. In order to solve the above two problems, this paper made some improvements on the construction of fusion features and genetic evolution algorithm in GERNNC model, and proposed an improved genetic evolution random neural network cluster (IGERNNC) model. The IGERNNC model uses mutual information correlation analysis method to combine resting-state functional magnetic resonance imaging data with single nucleotide polymorphism data for the construction of fusion features. Based on the traditional genetic evolution algorithm, elite retention strategy and large variation genetic algorithm are added to avoid the model falling into the local optimal solution. Through multiple independent experimental comparisons, the IGERNNC model can more effectively identify AD patients and extract relevant pathogenic factors, which is expected to become an effective tool in the field of AD research.

Quantifying an individual's risk for common diseases is an important goal of precision health. The polygenic risk score (PRS), which aggregates multiple risk alleles of candidate diseases, has emerged as a standard approach for identifying high-risk individuals. Although several studies have been performed to benchmark the PRS calculation tools and assess their potential to guide future clinical applications, some issues remain to be further investigated, such as lacking (i) various simulated data with different genetic effects; (ii) evaluation of machine learning models and (iii) evaluation on multiple ancestries studies. In this study, we systematically validated and compared 13 statistical methods, 5 machine learning models and 2 ensemble models using simulated data with additive and genetic interaction models, 22 common diseases with internal training sets, 4 common diseases with external summary statistics and 3 common diseases for trans-ancestry studies in UK Biobank. The statistical methods were better in simulated data from additive models and machine learning models have edges for data that include genetic interactions. Ensemble models are generally the best choice by integrating various statistical methods. LDpred2 outperformed the other standalone tools, whereas PRS-CS, lassosum and DBSLMM showed comparable performance. We also identified that disease heritability strongly affected the predictive performance of all methods. Both the number and effect sizes of risk SNPs are important; and sample size strongly influences the performance of all methods. For the trans-ancestry studies, we found that the performance of most methods became worse when training and testing sets were from different populations.

With the improvement of single-cell measurement techniques, there is a growing awareness that individual differences exist among cells, and protein expression distribution can vary across cells in the same tissue or cell line. Pinpointing the protein subcellular locations in single cells is crucial for mapping functional specificity of proteins and studying related diseases. Currently, research about single-cell protein location is still in its infancy, and most studies and databases do not annotate proteins at the cell level. For example, in the human protein atlas database, an immunofluorescence image stained for a particular protein shows multiple cells, but the subcellular location annotation is for the whole image, ignoring intercellular difference. In this study, we used large-scale immunofluorescence images and image-level subcellular locations to develop a deep-learning-based pipeline that could accurately recognize protein localizations in single cells. The pipeline consisted of two deep learning models, i.e. an image-based model and a cell-based model. The former used a multi-instance learning framework to comprehensively model protein distribution in multiple cells in each image, and could give both image-level and cell-level predictions. The latter firstly used clustering and heuristics algorithms to assign pseudo-labels of subcellular locations to the segmented cell images, and then used the pseudo-labels to train a classification model. Finally, the image-based model was fused with the cell-based model at the decision level to obtain the final ensemble model for single-cell prediction. Our experimental results showed that the ensemble model could achieve higher accuracy and robustness on independent test sets than state-of-the-art methods.

Synergistic drug combinations can improve the therapeutic effect and reduce the drug dosage to avoid toxicity. In previous years, an in vitro approach was utilized to screen synergistic drug combinations. However, the in vitro method is time-consuming and expensive. With the rapid growth of high-throughput data, computational methods are becoming efficient tools to predict potential synergistic drug combinations. Considering the limitations of the previous computational methods, we developed a new model named Siamese Network and Random Matrix Projection for AntiCancer Drug Combination prediction (SNRMPACDC). Firstly, the Siamese convolutional network and random matrix projection were used to process the features of the two drugs into drug combination features. Then, the features of the cancer cell line were processed through the convolutional network. Finally, the processed features were integrated and input into the multi-layer perceptron network to get the predicted score. Compared with the traditional method of splicing drug features into drug combination features, SNRMPACDC improved the interpretability of drug combination features to a certain extent. In addition, the introduction of convolutional networks can better extract the potential information in the features. SNRMPACDC achieved the root mean-squared error of 15.01 and the Pearson correlation coefficient of 0.75 in 5-fold cross-validation of regression prediction for response data. In addition, SNRMPACDC achieved the AUC of 0.91 ± 0.03 and the AUPR of 0.62 ± 0.05 in 5-fold cross-validation of classification prediction of synergistic or not. These results are almost better than all the previous models. SNRMPACDC would be an effective approach to infer potential anticancer synergistic drug combinations.

The inference of large-scale gene regulatory networks is essential for understanding comprehensive interactions among genes. Most existing methods are limited to reconstructing networks with a few hundred nodes. Therefore, parallel computing paradigms must be leveraged to construct large networks. We propose a generic parallel framework that enables any existing method, without re-engineering, to infer large networks in parallel, guaranteeing quality output. The framework is tested on 15 inference methods (not limited to) employing in silico benchmarks and real-world large expression matrices, followed by qualitative and speedup assessment. The framework does not compromise the quality of the base serial inference method. We rank the candidate methods and use the top-performing method to infer an Alzheimer's Disease (AD) affected network from large expression profiles of a triple transgenic mouse model consisting of 45,101 genes. The resultant network is further explored to obtain hub genes that emerge functionally related to the disease. We partition the network into 41 modules and conduct pathway enrichment analysis, revealing that a good number of participating genes are collectively responsible for several brain disorders, including AD. Finally, we extract the interactions of a few known AD genes and observe that they are periphery genes connected to the network's hub genes. Availability: The R implementation of the framework is downloadable from https://github.com/Netralab/GenericParallelFramework.

Genetic similarity matrices are commonly used to assess population substructure (PS) in genetic studies. Through simulation studies and by the application to whole-genome sequencing (WGS) data, we evaluate the performance of three genetic similarity matrices: the unweighted and weighted Jaccard similarity matrices and the genetic relationship matrix. We describe different scenarios that can create numerical pitfalls and lead to incorrect conclusions in some instances. We consider scenarios in which PS is assessed based on loci that are located across the genome ('globally') and based on loci from a specific genomic region ('locally'). We also compare scenarios in which PS is evaluated based on loci from different minor allele frequency bins: common (>5%), low-frequency (5-0.5%) and rare (<0.5%) single-nucleotide variations (SNVs). Overall, we observe that all approaches provide the best clustering performance when computed based on rare SNVs. The performance of the similarity matrices is very similar for common and low-frequency variants, but for rare variants, the unweighted Jaccard matrix provides preferable clustering features. Based on visual inspection and in terms of standard clustering metrics, its clusters are the densest and the best separated in the principal component analysis of variants with rare SNVs compared with the other methods and different allele frequency cutoffs. In an application, we assessed the role of rare variants on local and global PS, using WGS data from multiethnic Alzheimer's disease data sets and European or East Asian populations from the 1000 Genome Project.

Creating synthetic lines is the standard mating mode for commercial pig production. Traditional mating performance was evaluated through a strictly designed cross-combination test at the 'breed level' to maximize the benefits of production. The Duroc-Landrace-Yorkshire (DLY) three-way crossbred production system became the most widely used breeding scheme for pigs. Here, we proposed an 'individual level' genomic mating procedure that can be applied to commercial pig production with efficient algorithms for estimating marker effects and for allocating the appropriate boar-sow pairs, which can be freely accessed to public in our developed HIBLUP software at https://www.hiblup.com/tutorials#genomic-mating. A total of 875 Duroc boars, 350 Landrace-Yorkshire sows and 3573 DLY pigs were used to carry out the genomic mating to assess the production benefits theoretically. The results showed that genomic mating significantly improved the performances of progeny across different traits compared with random mating, such as the feed conversion rate, days from 30 to 120 kg and eye muscle area could be improved by -0.12, -4.64 d and 2.65 cm2, respectively, which were consistent with the real experimental validations. Overall, our findings indicated that genomic mating is an effective strategy to improve the performances of progeny by maximizing their total genetic merit with consideration of both additive and dominant effects. Also, a herd of boars from a richer genetic source will increase the effectiveness of genomic mating further.

Anticancer peptides (ACPs) are the types of peptides that have been demonstrated to have anticancer activities. Using ACPs to prevent cancer could be a viable alternative to conventional cancer treatments because they are safer and display higher selectivity. Due to ACP identification being highly lab-limited, expensive and lengthy, a computational method is proposed to predict ACPs from sequence information in this study. The process includes the input of the peptide sequences, feature extraction in terms of ordinal encoding with positional information and handcrafted features, and finally feature selection. The whole model comprises of two modules, including deep learning and machine learning algorithms. The deep learning module contained two channels: bidirectional long short-term memory (BiLSTM) and convolutional neural network (CNN). Light Gradient Boosting Machine (LightGBM) was used in the machine learning module. Finally, this study voted the three models' classification results for the three paths resulting in the model ensemble layer. This study provides insights into ACP prediction utilizing a novel method and presented a promising performance. It used a benchmark dataset for further exploration and improvement compared with previous studies. Our final model has an accuracy of 0.7895, sensitivity of 0.8153 and specificity of 0.7676, and it was increased by at least 2% compared with the state-of-the-art studies in all metrics. Hence, this paper presents a novel method that can potentially predict ACPs more effectively and efficiently. The work and source codes are made available to the community of researchers and developers at https://github.com/khanhlee/acp-ope/.

Due to the increasing importance of graphs and graph streams in data representation in today's era, concept drift detection in graph streaming scenarios is more important than ever. Contributions to concept drift detection in graph streams are minimal and practically non-existent in the field of toxicology. This paper applied the discriminative subgraph-based drift detector (DSDD) to graph streams generated from real-world toxicology datasets. We used four toxicology datasets, each of which yielded two graph streams - one with abrupt drift points and one with gradual drift points. We used DSDD both with the standard minimum description length (MDL) heuristic and after replacing MDL with a much simpler heuristic SIZE (number of vertices + number of edges), and applied it to all generated graph streams containing abrupt drift points and gradual drift points for varying window sizes. Following that, we compared and analyzed the results. Finally, we applied a long short-term memory based graph stream classification model to all the generated streams and compared the difference in the performances obtained with and without detecting drift using DSDD. We believe that the results and analysis presented in this paper will provide insight into the task of concept drift detection in the toxicology domain and aid in the application of DSDD in a variety of scenarios.

We draw from the assumption that similarities between pathogens at both pathogen protein and host protein level, may provide the appropriate framework to identify and rank candidate drugs to be used against a specific pathogen. Vir2Drug is a drug repurposing tool that uses network-based approaches to identify and rank candidate drugs for a specific pathogen, combining information obtained from: (a) ranked pathogen-to-pathogen networks based on protein similarities between pathogens, (b) taxonomy distance between pathogens and (c) drugs targeting specific pathogen's and host proteins. The underlying pathogen networks are used to screen drugs by means of specific methodologies that account for either the host or pathogen's protein targets. Vir2Drug is a useful and yet informative tool for drug repurposing against known or unknown pathogens especially in periods where the emergence for repurposed drugs plays significant role in handling viral outbreaks, until reaching a vaccine. The web tool is available at: https://bioinformatics.cing.ac.cy/vir2drug, https://vir2drug.cing-big.hpcf.cyi.ac.cy.

Longitudinal clonal tracking studies based on high-throughput sequencing technologies supported safety and long-term efficacy and unraveled hematopoietic reconstitution in many gene therapy applications with unprecedented resolution. However, monitoring patients over a decade-long follow-up entails a constant increase of large data volume with the emergence of critical computational challenges, unfortunately not addressed by currently available tools. Here we present ISAnalytics, a new R package for comprehensive and high-throughput clonal tracking studies using vector integration sites as markers of cellular identity. Once identified the clones externally from ISAnalytics and imported in the package, a wide range of implemented functionalities are available to users for assessing the safety and long-term efficacy of the treatment, here described in a clinical trial use case for Hurler disease, and for supporting hematopoietic stem cell biology in vivo with longitudinal analysis of clones over time, proliferation and differentiation. ISAnalytics is conceived to be metadata-driven, enabling users to focus on biological questions and hypotheses rather than on computational aspects. ISAnalytics can be fully integrated within laboratory workflows and standard procedures. Moreover, ISAnalytics is designed with efficient and scalable data structures, benchmarked with previous methods, and grants reproducibility and full analytical control through interactive web-reports and a module with Shiny interface. The implemented functionalities are flexible for all viral vector-based clonal tracking applications as well as genetic barcoding or cancer immunotherapies.

Spatially resolved transcriptomics technologies enable comprehensive measurement of gene expression patterns in the context of intact tissues. However, existing technologies suffer from either low resolution or shallow sequencing depth. Here, we present DIST, a deep learning-based method that imputes the gene expression profiles on unmeasured locations and enhances the gene expression for both original measured spots and imputed spots by self-supervised learning and transfer learning. We evaluate the performance of DIST for imputation, clustering, differential expression analysis and functional enrichment analysis. The results show that DIST can impute the gene expression accurately, enhance the gene expression for low-quality data, help detect more biological meaningful differentially expressed genes and pathways, therefore allow for deeper insights into the biological processes.

In longitudinal studies variables are measured repeatedly over time, leading to clustered and correlated observations. If the goal of the study is to develop prediction models, machine learning approaches such as the powerful random forest (RF) are often promising alternatives to standard statistical methods, especially in the context of high-dimensional data. In this paper, we review extensions of the standard RF method for the purpose of longitudinal data analysis. Extension methods are categorized according to the data structures for which they are designed. We consider both univariate and multivariate response longitudinal data and further categorize the repeated measurements according to whether the time effect is relevant. Even though most extensions are proposed for low-dimensional data, some can be applied to high-dimensional data. Information of available software implementations of the reviewed extensions is also given. We conclude with discussions on the limitations of our review and some future research directions.

Gene regulatory networks govern complex gene expression programs in various biological phenomena, including embryonic development, cell fate decisions and oncogenesis. Single-cell techniques are increasingly being used to study gene expression, providing higher resolution than traditional approaches. However, inferring a comprehensive gene regulatory network across different cell types remains a challenge. Here, we propose to construct context-dependent gene regulatory networks (CDGRNs) from single-cell RNA sequencing data utilizing both spliced and unspliced transcript expression levels. A gene regulatory network is decomposed into subnetworks corresponding to different transcriptomic contexts. Each subnetwork comprises the consensus active regulation pairs of transcription factors and their target genes shared by a group of cells, inferred by a Gaussian mixture model. We find that the union of gene regulation pairs in all contexts is sufficient to reconstruct differentiation trajectories. Functions specific to the cell cycle, cell differentiation or tissue-specific functions are enriched throughout the developmental process in each context. Surprisingly, we also observe that the network entropy of CDGRNs decreases along differentiation trajectories, indicating directionality in differentiation. Overall, CDGRN allows us to establish the connection between gene regulation at the molecular level and cell differentiation at the macroscopic level.

Microbial communities are highly dynamic and sensitive to changes in the environment. Thus, microbiome data are highly susceptible to batch effects, defined as sources of unwanted variation that are not related to and obscure any factors of interest. Existing batch effect correction methods have been primarily developed for gene expression data. As such, they do not consider the inherent characteristics of microbiome data, including zero inflation, overdispersion and correlation between variables. We introduce new multivariate and non-parametric batch effect correction methods based on Partial Least Squares Discriminant Analysis (PLSDA). PLSDA-batch first estimates treatment and batch variation with latent components, then subtracts batch-associated components from the data. The resulting batch-effect-corrected data can then be input in any downstream statistical analysis. Two variants are proposed to handle unbalanced batch x treatment designs and to avoid overfitting when estimating the components via variable selection. We compare our approaches with popular methods managing batch effects, namely, removeBatchEffect, ComBat and Surrogate Variable Analysis, in simulated and three case studies using various visual and numerical assessments. We show that our three methods lead to competitive performance in removing batch variation while preserving treatment variation, especially for unbalanced batch $\times $ treatment designs. Our downstream analyses show selections of biologically relevant taxa. This work demonstrates that batch effect correction methods can improve microbiome research outputs. Reproducible code and vignettes are available on GitHub.

Lysine glutarylation (Kglu) is a newly discovered post-translational modification of proteins with important roles in mitochondrial functions, oxidative damage, etc. The established biological experimental methods to identify glutarylation sites are often time-consuming and costly. Therefore, there is an urgent need to develop computational methods for efficient and accurate identification of glutarylation sites. Most of the existing computational methods only utilize handcrafted features to construct the prediction model and do not consider the positive impact of the pre-trained protein language model on the prediction performance. Based on this, we develop an ensemble deep-learning predictor Deepro-Glu that combines convolutional neural network and bidirectional long short-term memory network using the deep learning features and traditional handcrafted features to predict lysine glutaryation sites. The deep learning features are generated from the pre-trained protein language model called ProtBert, and the handcrafted features consist of sequence-based features, physicochemical property-based features and evolution information-based features. Furthermore, the attention mechanism is used to efficiently integrate the deep learning features and the handcrafted features by learning the appropriate attention weights. 10-fold cross-validation and independent tests demonstrate that Deepro-Glu achieves competitive or superior performance than the state-of-the-art methods. The source codes and data are publicly available at https://github.com/xwanggroup/Deepro-Glu.

DNA-methylation alterations are common in cancer and display unique characteristics that make them ideal markers for tumor quantification and classification. Here we present MIMESIS, a computational framework exploiting minimal DNA-methylation signatures composed by a few dozen informative DNA-methylation sites to quantify and classify tumor signals in tissue and cell-free DNA samples. Extensive analyses of multiple independent and heterogenous datasets including >7200 samples demonstrate the capability of MIMESIS to provide precise estimations of tumor content and to enable accurate classification of tumor type and molecular subtype. To assess our framework for clinical applications, we designed a MIMESIS-informed assay incorporating the minimal signatures for breast cancer. Using both artificial samples and clinical serial cell-free DNA samples from patients with metastatic breast cancer, we show that our approach provides accurate estimations of tumor content, sensitive detection of tumor signal and the ability to capture clinically relevant molecular subtype in patients' circulation. This study provides evidence that our extremely parsimonious approach can be used to develop cost-effective and highly scalable DNA-methylation assays that could support and facilitate the implementation of precision oncology in clinical practice.