Hepatocellular carcinoma (HCC) is one of the most common types of cancers and a global health challenge with a low early diagnosis rate and high mortality. The coagulation cascade plays an important role in the tumor immune microenvironment (TME) of HCC. In this study, based on the coagulation pathways collected from the KEGG database, two coagulation-related subtypes were distinguished in HCC patients. We demonstrated the distinct differences in immune characteristics and prognostic stratification between two coagulation-related subtypes. A coagulation-related risk score prognostic model was developed in the Cancer Genome Atlas (TCGA) cohort for risk stratification and prognosis prediction. The predictive values of the coagulation-related risk score in prognosis and immunotherapy were also verified in the TCGA and International Cancer Genome Consortium cohorts. A nomogram was also established to facilitate the clinical use of this risk score and verified its effectiveness using different approaches. Based on these results, we can conclude that there is an obvious correlation between the coagulation and the TME in HCC, and the risk score could serve as a robust prognostic biomarker, provide therapeutic benefits for chemotherapy and immunotherapy and may be helpful for clinical decision making in HCC patients.

Circular ribonucleic acids (RNAs) (circRNAs) are formed by covalently linking the downstream splice donor and the upstream splice acceptor. One of the most important functions of circRNAs is mainly exerted through binding RNA-binding proteins (RBPs). However, there is no efficient algorithm for identifying genome-wide circRNA-RBP interactions. Here, we developed a unique algorithm, circRIP, for identifying circRNA-RBP interactions from RNA immunoprecipitation sequencing (RIP-Seq) data. A simulation test demonstrated the sensitivity and specificity of circRIP. By applying circRIP, we identified 95 IGF2BP3-binding circRNAs based on the IGF2BP3 RIP-Seq dataset. We further identified 2823 and 1333 circRNAs binding to >100 RBPs in K562 and HepG2 cell lines, respectively, based on enhanced cross-linking immunoprecipitation (eCLIP) data, demonstrating the significance to survey the potential interactions between circRNAs and RBPs. In this study, we provide an accurate and sensitive tool, circRIP (https://github.com/bioinfolabwhu/circRIP), to systematically identify RBP and circRNA interactions from RIP-Seq and eCLIP data, which can significantly benefit the research community for the functional exploration of circRNAs.

Proteins are capable of highly specific interactions and are responsible for a wide range of functions, making them attractive in the pursuit of new therapeutic options. Previous studies focusing on overall geometry of protein-protein interfaces, however, concluded that PPI interfaces were generally flat. More recently, this idea has been challenged by their structural and thermodynamic characterisation, suggesting the existence of concave binding sites that are closer in character to traditional small-molecule binding sites, rather than exhibiting complete flatness. Here, we present a large-scale analysis of binding geometry and physicochemical properties of all protein-protein interfaces available in the Protein Data Bank. In this review, we provide a comprehensive overview of the protein-protein interface landscape, including evidence that even for overall larger, more flat interfaces that utilize discontinuous interacting regions, small and potentially druggable pockets are utilized at binding sites.

Complex biomedical data generated during clinical, omics and mechanism-based experiments have increasingly been exploited through cloud- and visualization-based data mining techniques. However, the scientific community still lacks an easy-to-use web service for the comprehensive visualization of biomedical data, particularly high-quality and publication-ready graphics that allow easy scaling and updatability according to user demands. Therefore, we propose a community-driven modern web service, Hiplot (https://hiplot.org), with concise and top-quality data visualization applications for the life sciences and biomedical fields. This web service permits users to conveniently and interactively complete a few specialized visualization tasks that previously could only be conducted by senior bioinformatics or biostatistics researchers. It covers most of the daily demands of biomedical researchers with its equipped 240+ biomedical data visualization functions, involving basic statistics, multi-omics, regression, clustering, dimensional reduction, meta-analysis, survival analysis, risk modelling, etc. Moreover, to improve the efficiency in use and development of plugins, we introduced some core advantages on the client-/server-side of the website, such as spreadsheet-based data importing, cross-platform command-line controller (Hctl), multi-user plumber workers, JavaScript Object Notation-based plugin system, easy data/parameters, results and errors reproduction and real-time updates mode. Meanwhile, using demo/real data sets and benchmark tests, we explored statistical parameters, cancer genomic landscapes, disease risk factors and the performance of website based on selected native plugins. The statistics of visits and user numbers could further reflect the potential impact of this web service on relevant fields. Thus, researchers devoted to life and data sciences would benefit from this emerging and free web service.

The microbiome is a complex and dynamic community of microorganisms that co-exist interdependently within an ecosystem, and interact with its host or environment. Longitudinal studies can capture temporal variation within the microbiome to gain mechanistic insights into microbial systems; however, current statistical methods are limited due to the complex and inherent features of the data. We have identified three analytical objectives in longitudinal microbial studies: (1) differential abundance over time and between sample groups, demographic factors or clinical variables of interest; (2) clustering of microorganisms evolving concomitantly across time and (3) network modelling to identify temporal relationships between microorganisms. This review explores the strengths and limitations of current methods to fulfill these objectives, compares different methods in simulation and case studies for objectives (1) and (2), and highlights opportunities for further methodological developments. R tutorials are provided to reproduce the analyses conducted in this review.

MOTIVATION: The control of the diffusion of diseases is a critical subject of a broad research area, which involves both clinical and political aspects. It makes wide use of computational tools, such as ordinary differential equations, stochastic simulation frameworks and graph theory, and interaction data, from molecular to social granularity levels, to model the ways diseases arise and spread. The coronavirus disease 2019 (COVID-19) is a perfect testbench example to show how these models may help avoid severe lockdown by suggesting, for instance, the best strategies of vaccine prioritization. RESULTS: Here, we focus on and discuss some graph-based epidemiological models and show how their use may significantly improve the disease spreading control. We offer some examples related to the recent COVID-19 pandemic and discuss how to generalize them to other diseases.

The network approach is quickly becoming a fundamental building block of computational methods aiming at elucidating the mechanism of action (MoA) and therapeutic effect of drugs. By modeling the effect of drugs and diseases on different biological networks, it is possible to better explain the interplay between disease perturbations and drug targets as well as how drug compounds induce favorable biological responses and/or adverse effects. Omics technologies have been extensively used to generate the data needed to study the mechanisms of action of drugs and diseases. These data are often exploited to define condition-specific networks and to study whether drugs can reverse disease perturbations. In this review, we describe network data mining algorithms that are commonly used to study drug's MoA and to improve our understanding of the basis of chronic diseases. These methods can support fundamental stages of the drug development process, including the identification of putative drug targets, the in silico screening of drug compounds and drug combinations for the treatment of diseases. We also discuss recent studies using biological and omics-driven networks to search for possible repurposed FDA-approved drug treatments for SARS-CoV-2 infections (COVID-19).

Patient similarity networks (PSNs), where patients are represented as nodes and their similarities as weighted edges, are being increasingly used in clinical research. These networks provide an insightful summary of the relationships among patients and can be exploited by inductive or transductive learning algorithms for the prediction of patient outcome, phenotype and disease risk. PSNs can also be easily visualized, thus offering a natural way to inspect complex heterogeneous patient data and providing some level of explainability of the predictions obtained by machine learning algorithms. The advent of high-throughput technologies, enabling us to acquire high-dimensional views of the same patients (e.g. omics data, laboratory data, imaging data), calls for the development of data fusion techniques for PSNs in order to leverage this rich heterogeneous information. In this article, we review existing methods for integrating multiple biomedical data views to construct PSNs, together with the different patient similarity measures that have been proposed. We also review methods that have appeared in the machine learning literature but have not yet been applied to PSNs, thus providing a resource to navigate the vast machine learning literature existing on this topic. In particular, we focus on methods that could be used to integrate very heterogeneous datasets, including multi-omics data as well as data derived from clinical information and medical imaging.

MOTIVATION: The study of the Human Virome remains challenging nowadays. Viral metagenomics, through high-throughput sequencing data, is the best choice for virus discovery. The metagenomics approach is culture-independent and sequence-independent, helping search for either known or novel viruses. Though it is estimated that more than 40% of the viruses found in metagenomics analysis are not recognizable, we decided to analyze several tools to identify and discover viruses in RNA-seq samples. RESULTS: We have analyzed eight Virus Tools for the identification of viruses in RNA-seq data. These tools were compared using a synthetic dataset of 30 viruses and a real one. Our analysis shows that no tool succeeds in recognizing all the viruses in the datasets. So we can conclude that each of these tools has pros and cons, and their choice depends on the application domain. AVAILABILITY: Synthetic data used through the review and raw results of their analysis can be found at https://zenodo.org/record/6426147. FASTQ files of real data can be found in GEO (https://www.ncbi.nlm.nih.gov/gds) or ENA (https://www.ebi.ac.uk/ena/browser/home). Raw results of their analysis can be downloaded from https://zenodo.org/record/6425917.

PIWI-interacting RNAs (piRNAs) are short 21-35 nucleotide molecules that comprise the largest class of non-coding RNAs and found in a large diversity of species including yeast, worms, flies, plants and mammals including humans. The most well-understood function of piRNAs is to monitor and protect the genome from transposons particularly in germline cells. Recent data suggest that piRNAs may have additional functions in somatic cells although they are expressed there in far lower abundance. Compared with microRNAs (miRNAs), piRNAs have more limited bioinformatics resources available. This review collates 39 piRNA specific and non-specific databases and bioinformatics resources, describes and compares their utility and attributes and provides an overview of their place in the field. In addition, we review 33 computational models based upon function: piRNA prediction, transposon element and mRNA-related piRNA prediction, cluster prediction, signature detection, target prediction and disease association. Based on the collection of databases and computational models, we identify trends and potential gaps in tool development. We further analyze the breadth and depth of piRNA data available in public sources, their contribution to specific human diseases, particularly in cancer and neurodegenerative conditions, and highlight a few specific piRNAs that appear to be associated with these diseases. This briefing presents the most recent and comprehensive mapping of piRNA bioinformatics resources including databases, models and tools for disease associations to date. Such a mapping should facilitate and stimulate further research on piRNAs.

The cost of drug development continues to rise and may be prohibitive in cases of unmet clinical need, particularly for rare diseases. Artificial intelligence-based methods are promising in their potential to discover new treatment options. The task of drug repurposing hypothesis generation is well-posed as a link prediction problem in a knowledge graph (KG) of interacting of drugs, proteins, genes and disease phenotypes. KGs derived from biomedical literature are semantically rich and up-to-date representations of scientific knowledge. Inference methods on scientific KGs can be confounded by unspecified contexts and contradictions. Extracting context enables incorporation of relevant pharmacokinetic and pharmacodynamic detail, such as tissue specificity of interactions. Contradictions in biomedical KGs may arise when contexts are omitted or due to contradicting research claims. In this review, we describe challenges to creating literature-scale representations of pharmacological knowledge and survey current approaches toward incorporating context and resolving contradictions.

Antibodies are versatile molecular binders with an established and growing role as therapeutics. Computational approaches to developing and designing these molecules are being increasingly used to complement traditional lab-based processes. Nowadays, in silico methods fill multiple elements of the discovery stage, such as characterizing antibody-antigen interactions and identifying developability liabilities. Recently, computational methods tackling such problems have begun to follow machine learning paradigms, in many cases deep learning specifically. This paradigm shift offers improvements in established areas such as structure or binding prediction and opens up new possibilities such as language-based modeling of antibody repertoires or machine-learning-based generation of novel sequences. In this review, we critically examine the recent developments in (deep) machine learning approaches to therapeutic antibody design with implications for fully computational antibody design.

Cyclin-dependent kinase (Cdk) proteins play crucial roles in the cell cycle progression and are thus attractive drug targets for therapy against such aberrant cell cycle processes as cancer. Since most of the available Cdk inhibitors target the highly conserved catalytic ATP pocket and their lack of specificity often lead to side effects, it is imperative to identify and characterize less conserved non-catalytic pockets capable of interfering with the kinase activity allosterically. However, a systematic analysis of these allosteric druggable pockets is still in its infancy. Here, we summarize the existing Cdk pockets and their selectivity. Then, we outline a network-based pocket prediction approach (NetPocket) and illustrate its utility for systematically identifying the allosteric druggable pockets with case studies. Finally, we discuss potential future directions and their challenges.

Identifying miRNA-disease associations is an important task for revealing pathogenic mechanism of complicated diseases. Different computational methods have been proposed. Although these methods obtained encouraging performance for detecting missing associations between known miRNAs and diseases, how to accurately predict associated diseases for new miRNAs is still a difficult task. In this regard, a ranking framework named idenMD-NRF is proposed for miRNA-disease association identification. idenMD-NRF treats the miRNA-disease association identification as an information retrieval task. Given a novel query miRNA, idenMD-NRF employs Learning to Rank algorithm to rank associated diseases based on high-level association features and various predictors. The experimental results on two independent test datasets indicate that idenMD-NRF is superior to other compared predictors. A user-friendly web server of idenMD-NRF predictor is freely available at http://bliulab.net/idenMD-NRF/.

The increase in the expectations of artificial intelligence (AI) technology has led to machine learning technology being actively used in the medical field. Non-negative matrix factorization (NMF) is a machine learning technique used for image analysis, speech recognition, and language processing; recently, it is being applied to medical research. Precision medicine, wherein important information is extracted from large-scale medical data to provide optimal medical care for every individual, is considered important in medical policies globally, and the application of machine learning techniques to this end is being handled in several ways. NMF is also introduced differently because of the characteristics of its algorithms. In this review, the importance of NMF in the field of medicine, with a focus on the field of oncology, is described by explaining the mathematical science of NMF and the characteristics of the algorithm, providing examples of how NMF can be used to establish precision medicine, and presenting the challenges of NMF. Finally, the direction regarding the effective use of NMF in the field of oncology is also discussed.

Multiplexed single-cell proteomes (SCPs) quantification by mass spectrometry greatly improves the SCP coverage. However, it still suffers from a low number of protein identifications and there is much room to boost proteins identification by computational methods. In this study, we present a novel framework DeepSCP, utilizing deep learning to boost SCP coverage. DeepSCP constructs a series of features of peptide-spectrum matches (PSMs) by predicting the retention time based on the multiple SCP sample sets and fragment ion intensities based on deep learning, and predicts PSM labels with an optimized-ensemble learning model. Evaluation of DeepSCP on public and in-house SCP datasets showed superior performances compared with other state-of-the-art methods. DeepSCP identified more confident peptides and proteins by controlling q-value at 0.01 using target-decoy competition method. As a convenient and low-cost computing framework, DeepSCP will help boost single-cell proteome identification and facilitate the future development and application of single-cell proteomics.

In recent work, researchers have paid considerable attention to the estimation of causal effects in observational studies with a large number of covariates, which makes the unconfoundedness assumption plausible. In this paper, we review propensity score (PS) methods developed in high-dimensional settings and broadly group them into model-based methods that extend models for prediction to causal inference and balance-based methods that combine covariate balancing constraints. We conducted systematic simulation experiments to evaluate these two types of methods, and studied whether the use of balancing constraints further improved estimation performance. Our comparison methods were post-double-selection (PDS), double-index PS (DiPS), outcome-adaptive LASSO (OAL), group LASSO and doubly robust estimation (GLiDeR), high-dimensional covariate balancing PS (hdCBPS), regularized calibrated estimators (RCAL) and approximate residual balancing method (balanceHD). For the four model-based methods, simulation studies showed that GLiDeR was the most stable approach, with high estimation accuracy and precision, followed by PDS, OAL and DiPS. For balance-based methods, hdCBPS performed similarly to GLiDeR in terms of accuracy, and outperformed balanceHD and RCAL. These findings imply that PS methods do not benefit appreciably from covariate balancing constraints in high-dimensional settings. In conclusion, we recommend the preferential use of GLiDeR and hdCBPS approaches for estimating causal effects in high-dimensional settings; however, further studies on the construction of valid confidence intervals are required.

The discovery of proper molecular signature from OMIC data is indispensable for determining biological state, physiological condition, disease etiology, and therapeutic response. However, the identified signature is reported to be highly inconsistent, and there is little overlap among the signatures identified from different biological datasets. Such inconsistency raises doubts about the reliability of reported signatures and significantly hampers its biological and clinical applications. Herein, an online tool, ConSIG, was constructed to realize consistent discovery of gene/protein signature from any uploaded transcriptomic/proteomic data. This tool is unique in a) integrating a novel strategy capable of significantly enhancing the consistency of signature discovery, b) determining the optimal signature by collective assessment, and c) confirming the biological relevance by enriching the disease/gene ontology. With the increasingly accumulated concerns about signature consistency and biological relevance, this online tool is expected to be used as an essential complement to other existing tools for OMIC-based signature discovery. ConSIG is freely accessible to all users without login requirement at https://idrblab.org/consig/.

The rate of biological data generation has increased dramatically in recent years, which has driven the importance of databases as a resource to guide innovation and the generation of biological insights. Given the complexity and scale of these databases, automatic data classification is often required. Biological data sets are often hierarchical in nature, with varying degrees of complexity, imposing different challenges to train, test and validate accurate and generalizable classification models. While some approaches to classify hierarchical data have been proposed, no guidelines regarding their utility, applicability and limitations have been explored or implemented. These include 'Local' approaches considering the hierarchy, building models per level or node, and 'Global' hierarchical classification, using a flat classification approach. To fill this gap, here we have systematically contrasted the performance of 'Local per Level' and 'Local per Node' approaches with a 'Global' approach applied to two different hierarchical datasets: BioLip and CATH. The results show how different components of hierarchical data sets, such as variation coefficient and prediction by depth, can guide the choice of appropriate classification schemes. Finally, we provide guidelines to support this process when embarking on a hierarchical classification task, which will help optimize computational resources and predictive performance.

With the development of high-throughput genotyping technology, single nucleotide polymorphism (SNP)-SNP interactions (SSIs) detection has become an essential way for understanding disease susceptibility. Various methods have been proposed to detect SSIs. However, given the disease complexity and bias of individual SSI detectors, these single-detector-based methods are generally unscalable for real genome-wide data and with unfavorable results. We propose a novel ensemble learning-based approach (ELSSI) that can significantly reduce the bias of individual detectors and their computational load. ELSSI randomly divides SNPs into different subsets and evaluates them by multi-type detectors in parallel. Particularly, ELSSI introduces a four-stage pipeline (generate, score, switch and filter) to iteratively generate new SNP combination subsets from SNP subsets, score the combination subset by individual detectors, switch high-score combinations to other detectors for re-scoring, then filter out combinations with low scores. This pipeline makes ELSSI able to detect high-order SSIs from large genome-wide datasets. Experimental results on various simulated and real genome-wide datasets show the superior efficacy of ELSSI to state-of-the-art methods in detecting SSIs, especially for high-order ones. ELSSI is applicable with moderate PCs on the Internet and flexible to assemble new detectors. The code of ELSSI is available at https://www.sdu-idea.cn/codes.php?name=ELSSI.

Medical Dialogue Information Extraction (MDIE) is a promising task for modern medical care systems, which greatly facilitates the development of many real-world applications such as electronic medical record generation, automatic disease diagnosis, etc. Recent methods have firstly achieved considerable performance in Chinese MDIE but still suffer from some inherent limitations, such as poor exploitation of the inter-dependencies in multiple utterances, weak discrimination of the hard samples. In this paper, we propose a contrastive multi-utterance inference (CMUI) method to address these issues. Specifically, we first use a type-aware encoder to provide an efficient encode mechanism toward different categories. Subsequently, we introduce a selective attention mechanism to explicitly capture the dependencies among utterances, which thus constructs a multi-utterance inference. Finally, a supervised contrastive learning approach is integrated into our framework to improve the recognition ability for the hard samples. Extensive experiments show that our model achieves state-of-the-art performance on a public benchmark Chinese-based dataset and delivers significant performance gain on MDIE as compared with baselines. Specifically, we outperform the state-of-the-art results in F1-score by 2.27%, 0.55% in Recall and 3.61% in Precision (The codes that support the findings of this study are openly available in CMUI at https://github.com/jc4357/CMUI.).

Erratum in Brief Bioinform. 2022 Nov 19;23(6):

The advances of single-cell DNA sequencing (scDNA-seq) enable us to characterize the genetic heterogeneity of cancer cells. However, the high noise and low coverage of scDNA-seq impede the estimation of copy number variations (CNVs). In addition, existing tools suffer from intensive execution time and often fail on large datasets. Here, we propose SeCNV, an efficient method that leverages structural entropy, to profile the copy numbers. SeCNV adopts a local Gaussian kernel to construct a matrix, depth congruent map (DCM), capturing the similarities between any two bins along the genome. Then, SeCNV partitions the genome into segments by minimizing the structural entropy from the DCM. With the partition, SeCNV estimates the copy numbers within each segment for cells. We simulate nine datasets with various breakpoint distributions and amplitudes of noise to benchmark SeCNV. SeCNV achieves a robust performance, i.e. the F1-scores are higher than 0.95 for breakpoint detections, significantly outperforming state-of-the-art methods. SeCNV successfully processes large datasets (>50 000 cells) within 4 min, while other tools fail to finish within the time limit, i.e. 120 h. We apply SeCNV to single-nucleus sequencing datasets from two breast cancer patients and acoustic cell tagmentation sequencing datasets from eight breast cancer patients. SeCNV successfully reproduces the distinct subclones and infers tumor heterogeneity. SeCNV is available at https://github.com/deepomicslab/SeCNV.

Internal validation is the most popular evaluation strategy used for drug-target predictive models. The simple random shuffling in the cross-validation, however, is not always ideal to handle large, diverse and copious datasets as it could potentially introduce bias. Hence, these predictive models cannot be comprehensively evaluated to provide insight into their general performance on a variety of use-cases (e.g. permutations of different levels of connectiveness and categories in drug and target space, as well as validations based on different data sources). In this work, we introduce a benchmark, BETA, that aims to address this gap by (i) providing an extensive multipartite network consisting of 0.97 million biomedical concepts and 8.5 million associations, in addition to 62 million drug-drug and protein-protein similarities and (ii) presenting evaluation strategies that reflect seven cases (i.e. general, screening with different connectivity, target and drug screening based on categories, searching for specific drugs and targets and drug repurposing for specific diseases), a total of seven Tests (consisting of 344 Tasks in total) across multiple sampling and validation strategies. Six state-of-the-art methods covering two broad input data types (chemical structure- and gene sequence-based and network-based) were tested across all the developed Tasks. The best-worst performing cases have been analyzed to demonstrate the ability of the proposed benchmark to identify limitations of the tested methods for running over the benchmark tasks. The results highlight BETA as a benchmark in the selection of computational strategies for drug repurposing and target discovery.

MOTIVATION: In contrast to messenger RNAs, the function of the wide range of existing long noncoding RNAs (lncRNAs) largely depends on their structure, which determines interactions with partner molecules. Thus, the determination or prediction of the secondary structure of lncRNAs is critical to uncover their function. Classical approaches for predicting RNA secondary structure have been based on dynamic programming and thermodynamic calculations. In the last 4 years, a growing number of machine learning (ML)-based models, including deep learning (DL), have achieved breakthrough performance in structure prediction of biomolecules such as proteins and have outperformed classical methods in short transcripts folding. Nevertheless, the accurate prediction for lncRNA still remains far from being effectively solved. Notably, the myriad of new proposals has not been systematically and experimentally evaluated. RESULTS: In this work, we compare the performance of the classical methods as well as the most recently proposed approaches for secondary structure prediction of RNA sequences using a unified and consistent experimental setup. We use the publicly available structural profiles for 3023 yeast RNA sequences, and a novel benchmark of well-characterized lncRNA structures from different species. Moreover, we propose a novel metric to assess the predictive performance of methods, exclusively based on the chemical probing data commonly used for profiling RNA structures, avoiding any potential bias incorporated by computational predictions when using dot-bracket references. Our results provide a comprehensive comparative assessment of existing methodologies, and a novel and public benchmark resource to aid in the development and comparison of future approaches. AVAILABILITY: Full source code and benchmark datasets are available at: https://github.com/sinc-lab/lncRNA-folding. CONTACT: lbugnon@sinc.unl.edu.ar.

MOTIVATION: Bacteriophages are viruses infecting bacteria. Being key players in microbial communities, they can regulate the composition/function of microbiome by infecting their bacterial hosts and mediating gene transfer. Recently, metagenomic sequencing, which can sequence all genetic materials from various microbiome, has become a popular means for new phage discovery. However, accurate and comprehensive detection of phages from the metagenomic data remains difficult. High diversity/abundance, and limited reference genomes pose major challenges for recruiting phage fragments from metagenomic data. Existing alignment-based or learning-based models have either low recall or precision on metagenomic data. RESULTS: In this work, we adopt the state-of-the-art language model, Transformer, to conduct contextual embedding for phage contigs. By constructing a protein-cluster vocabulary, we can feed both the protein composition and the proteins' positions from each contig into the Transformer. The Transformer can learn the protein organization and associations using the self-attention mechanism and predicts the label for test contigs. We rigorously tested our developed tool named PhaMer on multiple datasets with increasing difficulty, including quality RefSeq genomes, short contigs, simulated metagenomic data, mock metagenomic data and the public IMG/VR dataset. All the experimental results show that PhaMer outperforms the state-of-the-art tools. In the real metagenomic data experiment, PhaMer improves the F1-score of phage detection by 27%.

In recent years there has been a widespread interest in researching biomarkers of aging that could predict physiological vulnerability better than chronological age. Aging, in fact, is one of the most relevant risk factors for a wide range of maladies, and molecular surrogates of this phenotype could enable better patients stratification. Among the most promising of such biomarkers is DNA methylation-based biological age. Given the potential and variety of computational implementations (epigenetic clocks), we here present a systematic review of such clocks. Furthermore, we provide a large-scale performance comparison across different tissues and diseases in terms of age prediction accuracy and age acceleration, a measure of deviance from physiology. Our analysis offers both a state-of-the-art overview of the computational techniques developed so far and a heterogeneous picture of performances, which can be helpful in orienting future research.

Graph neural networks (GNNs) are the most promising deep learning models that can revolutionize non-Euclidean data analysis. However, their full potential is severely curtailed by poorly represented molecular graphs and features. Here, we propose a multiphysical graph neural network (MP-GNN) model based on the developed multiphysical molecular graph representation and featurization. All kinds of molecular interactions, between different atom types and at different scales, are systematically represented by a series of scale-specific and element-specific graphs with distance-related node features. From these graphs, graph convolution network (GCN) models are constructed with specially designed weight-sharing architectures. Base learners are constructed from GCN models from different elements at different scales, and further consolidated together using both one-scale and multi-scale ensemble learning schemes. Our MP-GNN has two distinct properties. First, our MP-GNN incorporates multiscale interactions using more than one molecular graph. Atomic interactions from various different scales are not modeled by one specific graph (as in traditional GNNs), instead they are represented by a series of graphs at different scales. Second, it is free from the complicated feature generation process as in conventional GNN methods. In our MP-GNN, various atom interactions are embedded into element-specific graph representations with only distance-related node features. A unique GNN architecture is designed to incorporate all the information into a consolidated model. Our MP-GNN has been extensively validated on the widely used benchmark test datasets from PDBbind, including PDBbind-v2007, PDBbind-v2013 and PDBbind-v2016. Our model can outperform all existing models as far as we know. Further, our MP-GNN is used in coronavirus disease 2019 drug design. Based on a dataset with 185 complexes of inhibitors for severe acute respiratory syndrome coronavirus (SARS-CoV/SARS-CoV-2), we evaluate their binding affinities using our MP-GNN. It has been found that our MP-GNN is of high accuracy. This demonstrates the great potential of our MP-GNN for the screening of potential drugs for SARS-CoV-2. Availability: The Multiphysical graph neural network (MP-GNN) model can be found in https://github.com/Alibaba-DAMO-DrugAI/MGNN. Additional data or code will be available upon reasonable request.

In 2014, the Immune Epitope Database automated benchmark was created to compare the performance of the MHC class I binding predictors. However, this is not a straightforward process due to the different and non-standardized outputs of the methods. Additionally, some methods are more restrictive regarding the HLA alleles and epitope sizes for which they predict binding affinities, while others are more comprehensive. To address how these problems impacted the ranking of the predictors, we developed an approach to assess the reliability of different metrics. We found that using percentile-ranked results improved the stability of the ranks and allowed the predictors to be reliably ranked despite not being evaluated on the same data. We also found that given the rate new data are incorporated into the benchmark, a new method must wait for at least 4 years to be ranked against the pre-existing methods. The best-performing tools with statistically indistinguishable scores in this benchmark were NetMHCcons, NetMHCpan4.0, ANN3.4, NetMHCpan3.0 and NetMHCpan2.8. The results of this study will be used to improve the evaluation and display of benchmark performance. We highly encourage anyone working on MHC binding predictions to participate in this benchmark to get an unbiased evaluation of their predictors.

Oxidative stress is known to be involved in and possibly a key driver of the development of numerous chronic diseases, including cancer. It is highly desired to have a capability to reliably estimate the level of intracellular oxidative stress as it can help to identify functional changes and disease phenotypes associated with such a stress, but the problem proves to be very challenging. We present a novel computational model for quantitatively estimating the level of oxidative stress in tissues and cells based on their transcriptomic data. The model consists of (i) three sets of marker genes found to be associated with the production of oxidizing molecules, the activated antioxidation programs and the intracellular stress attributed to oxidation, respectively; (ii) three polynomial functions defined over the expression levels of the three gene sets are developed aimed to capture the total oxidizing power, the activated antioxidation capacity and the oxidative stress level, respectively, with their detailed parameters estimated by solving an optimization problem and (iii) the optimization problem is so formulated to capture the relevant known insights such as the oxidative stress level generally goes up from normal to chronic diseases and then to cancer tissues. Systematic assessments on independent datasets indicate that the trained predictor is highly reliable and numerous insights are made based on its application results to samples in the TCGA, GTEx and GEO databases.

The generation of candidate hit molecules with the potential to be used in cancer treatment is a challenging task. In this context, computational methods based on deep learning have been employed to improve in silico drug design methodologies. Nonetheless, the applied strategies have focused solely on the chemical aspect of the generation of compounds, disregarding the likely biological consequences for the organism's dynamics. Herein, we propose a method to implement targeted molecular generation that employs biological information, namely, disease-associated gene expression data, to conduct the process of identifying interesting hits. When applied to the generation of USP7 putative inhibitors, the framework managed to generate promising compounds, with more than 90% of them containing drug-like properties and essential active groups for the interaction with the target. Hence, this work provides a novel and reliable method for generating new promising compounds focused on the biological context of the disease.

Fungal extracellular vesicles (EVs) are released during pathogenesis and are found to be an opportunistic infection in most cases. EVs are immunocompetent with their host and have paved the way for new biomedical approaches to drug delivery and the treatment of complex diseases including cancer. With computing and processing advancements, the rise of bioinformatics tools for the evaluation of various parameters involved in fungal EVs has blossomed. In this review, we have complied and explored the bioinformatics tools to analyze the host-pathogen interaction, toxicity, omics and pathogenesis with an array of specific tools that have depicted the ability of EVs as vector/carrier for therapeutic agents and as a potential theme for immunotherapy. We have also discussed the generation and pathways involved in the production, transport, pathogenic action and immunological interactions of EVs in the host system. The incorporation of network pharmacology approaches has been discussed regarding fungal pathogens and their significance in drug discovery. To represent the overview, we have presented and demonstrated an in silico study model to portray the human Cryptococcal interactions.

Current tailored-therapy efforts in cancer are largely focused on a small number of highly recurrently mutated driver genes but therapeutic targeting of these oncogenes remains challenging. However, the vast number of genes mutated infrequently across cancers has received less attention, in part, due to a lack of understanding of their biological significance. We present SYSMut, an extendable systems biology platform that can robustly infer the biologic consequences of somatic mutations by integrating routine multiomics profiles in primary tumors. We establish SYSMut's improved performance vis-à-vis state-of-the-art driver gene identification methodologies by recapitulating the functional impact of known driver genes, while additionally identifying novel functionally impactful mutated genes across 29 cancers. Subsequent application of SYSMut on low-frequency gene mutations in head and neck squamous cell (HNSC) cancers, followed by molecular and pharmacogenetic validation, revealed the lipidogenic network as a novel therapeutic vulnerability in aggressive HNSC cancers. SYSMut is thus a robust scalable framework that enables the discovery of new targetable avenues in cancer.

The interplay between protein and nucleic acid participates in diverse biological activities. Accurately identifying the interaction between protein and nucleic acid can strengthen the understanding of protein function. However, conventional methods are too time-consuming, and computational methods are type-agnostic predictions. We proposed an ensemble predictor termed TSNAPred and first used it to identify residues that bind to A-DNA, B-DNA, ssDNA, mRNA, tRNA and rRNA. TSNAPred combines LightGBM and capsule network, both learned on the feature derived from protein sequence. TSNAPred utilizes the sliding window technique to extract long-distance dependencies between residues and a weighted ensemble strategy to enhance the prediction performance. The results show that TSNAPred can effectively identify type-specific nucleic acid binding residues in our test set. What is more, it also can discriminate DNA-binding and RNA-binding residues, which has improved 5% to 10% on the AUC value compared with other state-of-the-art methods. The dataset and code of TSNAPred are available at: https://github.com/niewenjuan-csu/TSNAPred.

Computational prediction of multiple-type drug-drug interaction (DDI) helps reduce unexpected side effects in poly-drug treatments. Although existing computational approaches achieve inspiring results, they ignore to study which local structures of drugs cause DDIs, and their interpretability is still weak. In this paper, by supposing that the interactions between two given drugs are caused by their local chemical structures (substructures) and their DDI types are determined by the linkages between different substructure sets, we design a novel Substructure-aware Tensor Neural Network model for DDI prediction (STNN-DDI). The proposed model learns a 3-D tensor of $\langle $  substructure, substructure, interaction type  $\rangle $ triplets, which characterizes a substructure-substructure interaction (SSI) space. According to a list of predefined substructures with specific chemical meanings, the mapping of drugs into this SSI space enables STNN-DDI to perform the multiple-type DDI prediction in both transductive and inductive scenarios in a unified form with an explicable manner. The comparison with deep learning-based state-of-the-art baselines demonstrates the superiority of STNN-DDI with the significant improvement of AUC, AUPR, Accuracy and Precision. More importantly, case studies illustrate its interpretability by both revealing an important substructure pair across drugs regarding a DDI type of interest and uncovering interaction type-specific substructure pairs in a given DDI. In summary, STNN-DDI provides an effective approach to predicting DDIs as well as explaining the interaction mechanisms among drugs. Source code is freely available at https://github.com/zsy-9/STNN-DDI.

Developments of single-cell RNA sequencing (scRNA-seq) technologies have enabled biological discoveries at the single-cell resolution with high throughput. However, large scRNA-seq datasets always suffer from massive technical noises, including batch effects and dropouts, and the dropout is often shown to be batch-dependent. Most existing methods only address one of the problems, and we show that the popularly used methods failed in trading off batch effect correction and dropout imputation. Here, inspired by the idea of causal inference, we propose a novel propensity score matching method for scRNA-seq data (scPSM) by borrowing information and taking the weighted average from similar cells in the deep sequenced batch, which simultaneously removes the batch effect, imputes dropout and denoises data in the entire gene expression space. The proposed method is testified on two simulation datasets and a variety of real scRNA-seq datasets, and the results show that scPSM is superior to other state-of-the-art methods. First, scPSM improves clustering accuracy and mixes cells of the same type, suggesting its ability to keep cell type separation while correcting for batch. Besides, using the scPSM-integrated data as input yields results free of batch effects or dropouts in the differential expression analysis. Moreover, scPSM not only achieves ideal denoising but also preserves real biological structure for downstream gene-based analyses. Furthermore, scPSM is robust to hyperparameters and small datasets with a few cells but enormous genes. Comprehensive evaluations demonstrate that scPSM jointly provides desirable batch effect correction, imputation and denoising for recovering the biologically meaningful expression in scRNA-seq data.

Nanopore sequencing is an emerging technology that reads DNA by utilizing a unique method of detecting nucleic acid sequences and identifies the various chemical modifications they carry. Deep learning has increased in popularity as a useful technique to solve many complex computational tasks. 'Adaptive sequencing' is an implementation of selective sequencing, intended for use on the nanopore sequencing platform. In this study, we demonstrated an alternative method of software-based selective sequencing that is performed in real time by combining nanopore sequencing and deep learning. Our results showed the feasibility of using deep learning for classifying signals from only the first 200 nucleotides in a raw nanopore sequencing signal format. This was further demonstrated by comparing the accuracy of our deep learning classification model across data from several human cell lines and other eukaryotic organisms. We used custom deep learning models and a script that utilizes a 'Read Until' framework to target mitochondrial molecules in real time from a human cell line sample. This achieved a significant separation and enrichment ability of 2.3-fold. In a series of very short sequencing experiments (10, 30 and 120 min), we identified genomic and mitochondrial reads with accuracy above 90%, although mitochondrial DNA comprised only 0.1% of the total input material. The uniqueness of our method is the ability to distinguish two groups of DNA even without a labeled reference. This contrasts with studies that required a well-defined reference, whether of a DNA sequence or of another type of representation. Additionally, our method showed higher correlation to the theoretically possible enrichment factor, compared with other published methods. We believe that our results will lay the foundation for rapid and selective sequencing using nanopore technology and will pave the approach for clinical applications that use nanopore sequencing data.

Rapid progresses in RNA-Seq and computational methods have assisted in quantifying A-to-I RNA editing and altered RNA editing sites have been widely observed in various diseases. Nevertheless, functional characterization of the altered RNA editing sites still remains a challenge. Here, we developed perturbations of RNA editing sites (PRES; http://bio-bigdata.hrbmu.edu.cn/PRES/) as the webserver for decoding functional perturbations of RNA editing sites based on editome profiling. After uploading an editome profile among samples of different groups, PRES will first annotate the editing sites to various genomic elements and detect differential editing sites under the user-selected method and thresholds. Next, the downstream functional perturbations of differential editing sites will be characterized from gain or loss miRNA/RNA binding protein regulation, RNA and protein structure changes, and the perturbed biological pathways. A prioritization module was developed to rank genes based on their functional consequences of RNA editing events. PRES provides user-friendly functionalities, ultra-efficient calculation, intuitive table and figure visualization interface to display the annotated RNA editing events, filtering options and elaborate application notebooks. We anticipate PRES will provide an opportunity for better understanding the regulatory mechanisms of RNA editing in human complex diseases.

Disease pathogenesis is always a major topic in biomedical research. With the exponential growth of biomedical information, drug effect analysis for specific phenotypes has shown great promise in uncovering disease-associated pathways. However, this method has only been applied to a limited number of drugs. Here, we extracted the data of 4634 diseases, 3671 drugs, 112 809 disease-drug associations and 81 527 drug-gene associations by text mining of 29 168 919 publications. On this basis, we proposed a 'Drug Set Enrichment Analysis by Text Mining (DSEATM)' pipeline and applied it to 3250 diseases, which outperformed the state-of-the-art method. Furthermore, diseases pathways enriched by DSEATM were similar to those obtained using the TCGA cancer RNA-seq differentially expressed genes. In addition, the drug number, which showed a remarkable positive correlation of 0.73 with the AUC, plays a determining role in the performance of DSEATM. Taken together, DSEATM is an auspicious and accurate disease research tool that offers fresh insights.

Predicting the drug-target interaction is crucial for drug discovery as well as drug repurposing. Machine learning is commonly used in drug-target affinity (DTA) problem. However, the machine learning model faces the cold-start problem where the model performance drops when predicting the interaction of a novel drug or target. Previous works try to solve the cold start problem by learning the drug or target representation using unsupervised learning. While the drug or target representation can be learned in an unsupervised manner, it still lacks the interaction information, which is critical in drug-target interaction. To incorporate the interaction information into the drug and protein interaction, we proposed using transfer learning from chemical-chemical interaction (CCI) and protein-protein interaction (PPI) task to drug-target interaction task. The representation learned by CCI and PPI tasks can be transferred smoothly to the DTA task due to the similar nature of the tasks. The result on the DTA datasets shows that our proposed method has advantages compared to other pre-training methods in the DTA task.

Candidate compounds with high binding affinities toward a target protein are likely to be developed as drugs. Deep neural networks (DNNs) have attracted increasing attention for drug-target affinity (DTA) estimation owning to their efficiency. However, the negative impact of batch effects caused by measure metrics, system technologies and other assay information is seldom discussed when training a DNN model for DTA. Suffering from the data deviation caused by batch effects, the DNN models can only be trained on a small amount of 'clean' data. Thus, it is challenging for them to provide precise and consistent estimations. We design a batch-sensitive training framework, namely BatchDTA, to train the DNN models. BatchDTA implicitly aligns multiple batches toward the same protein through learning the orders of candidate compounds with respect to the batches, alleviating the impact of the batch effects on the DNN models. Extensive experiments demonstrate that BatchDTA facilitates four mainstream DNN models to enhance the ability and robustness on multiple DTA datasets (BindingDB, Davis and KIBA). The average concordance index of the DNN models achieves a relative improvement of 4.0%. The case study reveals that BatchDTA can successfully learn the ranking orders of the compounds from multiple batches. In addition, BatchDTA can also be applied to the fused data collected from multiple sources to achieve further improvement.

Drug-induced liver injury (DILI) is one of the most significant concerns in medical practice but yet it still cannot be fully recapitulated with existing in vivo, in vitro and in silico approaches. To address this challenge, Chen et al. [ 1] developed a deep learning-based DILI prediction model based on chemical structure information alone. The reported model yielded an outstanding prediction performance (i.e. 0.958, 0.976, 0.935, 0.947, 0.926 and 0.913 for AUC, accuracy, recall, precision, F1-score and specificity, respectively, on a test set), far outperforming all publicly available and similar in silico DILI models. This extraordinary model performance is counter-intuitive to what we know about the underlying biology of DILI and the principles and hypothesis behind this type of in silico approach. In this Letter to the Editor, we raise awareness of several issues concerning data curation, model validation and comparison practices, and data and model reproducibility.

Viruses are ubiquitous in humans and various environments and continually mutate themselves. Identifying viruses in an environment without cultivation is challenging; however, promoting the screening of novel viruses and expanding the knowledge of viral space is essential. Homology-based methods that identify viruses using known viral genomes rely on sequence alignments, making it difficult to capture remote homologs of the known viruses. To accurately capture viral signals from metagenomic samples, models are needed to understand the patterns encoded in the viral genomes. In this study, we developed a hierarchical BERT model named ViBE to detect eukaryotic viruses from metagenome sequencing data and classify them at the order level. We pre-trained ViBE using read-like sequences generated from the virus reference genomes and derived three fine-tuned models that classify paired-end reads to orders for eukaryotic deoxyribonucleic acid viruses and eukaryotic ribonucleic acid viruses. ViBE achieved higher recall than state-of-the-art alignment-based methods while maintaining comparable precision. ViBE outperformed state-of-the-art alignment-free methods for all test cases. The performance of ViBE was also verified using real sequencing datasets, including the vaginal virome.

Computational models are often employed in systems biology to study the dynamic behaviours of complex systems. With the rise in the number of computational models, finding ways to improve the reusability of these models and their ability to reproduce virtual experiments becomes critical. Correct and effective model annotation in community-supported and standardised formats is necessary for this improvement. Here, we present recent efforts toward a common framework for annotated, accessible, reproducible and interoperable computational models in biology, and discuss key challenges of the field.

Single-cell RNA sequencing trades read-depth for dimensionality, often leading to loss of critical signaling gene information that is typically present in bulk data sets. We introduce DURIAN (Deconvolution and mUltitask-Regression-based ImputAtioN), an integrative method for recovery of gene expression in single-cell data. Through systematic benchmarking, we demonstrate the accuracy, robustness and empirical convergence of DURIAN using both synthetic and published data sets. We show that use of DURIAN improves single-cell clustering, low-dimensional embedding, and recovery of intercellular signaling networks. Our study resolves several inconsistent results of cell-cell communication analysis using single-cell or bulk data independently. The method has broad application in biomarker discovery and cell signaling analysis using single-cell transcriptomics data sets.

We construct a protein-protein interaction (PPI) targeted drug-likeness dataset and propose a deep molecular generative framework to generate novel drug-likeness molecules from the features of the seed compounds. This framework gains inspiration from published molecular generative models, uses the key features associated with PPI inhibitors as input and develops deep molecular generative models for de novo molecular design of PPI inhibitors. For the first time, quantitative estimation index for compounds targeting PPI was applied to the evaluation of the molecular generation model for de novo design of PPI-targeted compounds. Our results estimated that the generated molecules had better PPI-targeted drug-likeness and drug-likeness. Additionally, our model also exhibits comparable performance to other several state-of-the-art molecule generation models. The generated molecules share chemical space with iPPI-DB inhibitors as demonstrated by chemical space analysis. The peptide characterization-oriented design of PPI inhibitors and the ligand-based design of PPI inhibitors are explored. Finally, we recommend that this framework will be an important step forward for the de novo design of PPI-targeted therapeutics.

Even though several in silico tools are available for prediction of the phosphorylation sites for mammalian, yeast or plant proteins, currently no software is available for predicting phosphosites for Plasmodium proteins. However, the availability of significant amount of phospho-proteomics data during the last decade and advances in machine learning (ML) algorithms have opened up the opportunities for deciphering phosphorylation patterns of plasmodial system and developing ML-based phosphosite prediction tools for Plasmodium. We have developed Pf-Phospho, an ML-based method for prediction of phosphosites by training Random Forest classifiers using a large data set of 12 096 phosphosites of Plasmodium falciparum and Plasmodium bergei. Of the 12 096 known phosphosites, 75% of sites have been used for training/validation of the classifier, while remaining 25% have been used as completely unseen test data for blind testing. It is encouraging to note that Pf-Phospho can predict the kinase-independent phosphosites with 84% sensitivity, 75% specificity and 78% precision. In addition, it can also predict kinase-specific phosphosites for five plasmodial kinases-PfPKG, Plasmodium falciparum, PfPKA, PfPK7 and PbCDPK4 with high accuracy. Pf-Phospho (http://www.nii.ac.in/pfphospho.html) outperforms other widely used phosphosite prediction tools, which have been trained using mammalian phosphoproteome data. It also has been integrated with other widely used resources such as PlasmoDB, MPMP, Pfam and recently available ML-based predicted structures by AlphaFold2. Currently, Pf-phospho is the only bioinformatics resource available for ML-based prediction of phospho-signaling networks of Plasmodium and is a user-friendly platform for integrative analysis of phospho-signaling along with metabolic and protein-protein interaction networks.

DNA methylation analysis by sequencing is becoming increasingly popular, yielding methylomes at single-base pair and single-molecule resolution. It has tremendous potential for cell-type heterogeneity analysis using intrinsic read-level information. Although diverse deconvolution methods were developed to infer cell-type composition based on bulk sequencing-based methylomes, systematic evaluation has not been performed yet. Here, we thoroughly benchmark six previously published methods: Bayesian epiallele detection, DXM, PRISM, csmFinder+coMethy, ClubCpG and MethylPurify, together with two array-based methods, MeDeCom and Houseman, as a comparison group. Sequencing-based deconvolution methods consist of two main steps, informative region selection and cell-type composition estimation, thus each was individually assessed. With this elaborate evaluation, we aimed to establish which method achieves the highest performance in different scenarios of synthetic bulk samples. We found that cell-type deconvolution performance is influenced by different factors depending on the number of cell types within the mixture. Finally, we propose a best-practice deconvolution strategy for sequencing data and point out limitations that need to be handled. Array-based methods-both reference-based and reference-free-generally outperformed sequencing-based methods, despite the absence of read-level information. This implies that the current sequencing-based methods still struggle with correctly identifying cell-type-specific signals and eliminating confounding methylation patterns, which needs to be handled in future studies.

The poly(A) tail is a dynamic addition to the eukaryotic mRNA and the change in its length plays an essential role in regulating gene expression through affecting nuclear export, mRNA stability and translation. Only recently high-throughput sequencing strategies began to emerge for transcriptome-wide profiling of poly(A) tail length in diverse developmental stages and organisms. However, there is currently no easy-to-use and universal tool for measuring poly(A) tails in sequencing data from different sequencing protocols. Here we established PolyAtailor, a unified and efficient framework, for identifying and analyzing poly(A) tails from PacBio-based long reads or next generation short reads. PolyAtailor provides two core functions for measuring poly(A) tails, namely Tail_map and Tail_scan, which can be used for profiling tails with or without using a reference genome. Particularly, PolyAtailor can identify all potential tails in a read, providing users with detailed information such as tail position, tail length, tail sequence and tail type. Moreover, PolyAtailor integrates rich functions for poly(A) tail and poly(A) site analyses, such as differential poly(A) length analysis, poly(A) site identification and annotation, and statistics and visualization of base composition in tails. We compared PolyAtailor with three latest methods, FLAMAnalysis, FLEPSeq and PAIsoSeqAnalysis, using data from three sequencing protocols in HeLa samples and Arabidopsis. Results show that PolyAtailor is effective in measuring poly(A) tail length and detecting significance of differential poly(A) length, which achieves much higher sensitivity and accuracy than competing methods. PolyAtailor is available at https://github.com/BMILAB/PolyAtailor.

Despite the rapid development of sequencing technology, single-nucleotide polymorphism (SNP) arrays are still the most cost-effective genotyping solutions for large-scale genomic research and applications. Recent years have witnessed the rapid development of numerous genotyping platforms of different sizes and designs, but population-specific platforms are still lacking, especially for those in developing countries. SNP arrays designed for these countries should be cost-effective (small size), yet incorporate key information needed to associate genotypes with traits. A key design principle for most current platforms is to improve genome-wide imputation so that more SNPs not included in the array (imputed SNPs) can be predicted. However, current tag SNP selection methods mostly focus on imputation accuracy and coverage, but not the functional content of the array. It is those functional SNPs that are most likely associated with traits. Here, we propose LmTag, a novel method for tag SNP selection that not only improves imputation performance but also prioritizes highly functional SNP markers. We apply LmTag on a wide range of populations using both public and in-house whole-genome sequencing databases. Our results show that LmTag improved both functional marker prioritization and genome-wide imputation accuracy compared to existing methods. This novel approach could contribute to the next generation genotyping arrays that provide excellent imputation capability as well as facilitate array-based functional genetic studies. Such arrays are particularly suitable for under-represented populations in developing countries or non-model species, where little genomics data are available while investment in genome sequencing or high-density SNP arrays is limited. $\textrm{LmTag}$ is available at: https://github.com/datngu/LmTag.

The development of single-cell RNA-seq (scRNA-seq) technology allows researchers to characterize the cell types, states and transitions during dynamic biological processes at single-cell resolution. One of the critical tasks is to infer pseudo-time trajectory. However, the existence of transition cells in the intermediate state of complex biological processes poses a challenge for the trajectory inference. Here, we propose a new single-cell trajectory inference method based on transition entropy, named scTite, to identify transitional states and reconstruct cell trajectory from scRNA-seq data. Taking into account the continuity of cellular processes, we introduce a new metric called transition entropy to measure the uncertainty of a cell belonging to different cell clusters, and then identify cell states and transition cells. Specifically, we adopt different strategies to infer the trajectory for the identified cell states and transition cells, and combine them to obtain a detailed cell trajectory. For the identified cell clusters, we utilize the Wasserstein distance based on the probability distribution to calculate distance between clusters, and construct the minimum spanning tree. Meanwhile, we adopt the signaling entropy and partial correlation coefficient to determine transition paths, which contain a group of transition cells with the largest similarity. Then the transitional paths and the MST are combined to infer a refined cell trajectory. We apply scTite to four real scRNA-seq datasets and an integrated dataset, and conduct extensive performance comparison with nine existing trajectory inference methods. The experimental results demonstrate that the proposed method can reconstruct the cell trajectory more accurately than the compared algorithms. The scTite software package is available at https://github.com/dblab2022/scTite.

Finding direct dependencies between genetic pathways and diseases has been the target of multiple studies as it has many applications. However, due to cellular heterogeneity and limitations of the number of samples for bulk expression profiles, such studies have faced hurdles in the past. Here, we propose a method to perform single-cell expression-based inference of association between pathway, disease and cell-type (sci-PDC), which can help to understand their cause and effect and guide precision therapy. Our approach highlighted reliable relationships between a few diseases and pathways. Using the example of diabetes, we have demonstrated how sci-PDC helps in tracking variation of association between pathways and diseases with changes in age and species. The variation in pathways-disease associations in mice and humans revealed critical facts about the suitability of the mouse model for a few pathways in the context of diabetes. The coherence between results from our method and previous reports, including information about the drug target pathways, highlights its reliability for multidimensional utility.

Protein-DNA and protein-RNA interactions are involved in many biological activities. In the post-genome era, accurate identification of DNA- and RNA-binding residues in protein sequences is of great significance for studying protein functions and promoting new drug design and development. Therefore, some sequence-based computational methods have been proposed for identifying DNA- and RNA-binding residues. However, they failed to fully utilize the functional properties of residues, leading to limited prediction performance. In this paper, a sequence-based method iDRNA-ITF was proposed to incorporate the functional properties in residue representation by using an induction and transfer framework. The properties of nucleic acid-binding residues were induced by the nucleic acid-binding residue feature extraction network, and then transferred into the feature integration modules of the DNA-binding residue prediction network and the RNA-binding residue prediction network for the final prediction. Experimental results on four test sets demonstrate that iDRNA-ITF achieves the state-of-the-art performance, outperforming the other existing sequence-based methods. The webserver of iDRNA-ITF is freely available at http://bliulab.net/iDRNA-ITF.

Dynamic Bayesian networks (DBNs) can be used for the discovery of gene regulatory networks (GRNs) from time series gene expression data. Here, we suggest a strategy for learning DBNs from gene expression data by employing a Bayesian approach that is scalable to large networks and is targeted at learning models with high predictive accuracy. Our framework can be used to learn DBNs for multiple groups of samples and highlight differences and similarities in their GRNs. We learn these DBN models based on different structural and parametric assumptions and select the optimal model based on the cross-validated predictive accuracy. We show in simulation studies that our approach is better equipped to prevent overfitting than techniques used in previous studies. We applied the proposed DBN-based approach to two time series transcriptomic datasets from the Gene Expression Omnibus database, each comprising data from distinct phenotypic groups of the same tissue type. In the first case, we used DBNs to characterize responders and non-responders to anti-cancer therapy. In the second case, we compared normal to tumor cells of colorectal tissue. The classification accuracy reached by the DBN-based classifier for both datasets was higher than reported previously. For the colorectal cancer dataset, our analysis suggested that GRNs for cancer and normal tissues have a lot of differences, which are most pronounced in the neighborhoods of oncogenes and known cancer tissue markers. The identified differences in gene networks of cancer and normal cells may be used for the discovery of targeted therapies.

The COVID-19 pandemic caused several million deaths worldwide. Development of anti-coronavirus drugs is thus urgent. Unlike conventional non-peptide drugs, antiviral peptide drugs are highly specific, easy to synthesize and modify, and not highly susceptible to drug resistance. To reduce the time and expense involved in screening thousands of peptides and assaying their antiviral activity, computational predictors for identifying anti-coronavirus peptides (ACVPs) are needed. However, few experimentally verified ACVP samples are available, even though a relatively large number of antiviral peptides (AVPs) have been discovered. In this study, we attempted to predict ACVPs using an AVP dataset and a small collection of ACVPs. Using conventional features, a binary profile and a word-embedding word2vec (W2V), we systematically explored five different machine learning methods: Transformer, Convolutional Neural Network, bidirectional Long Short-Term Memory, Random Forest (RF) and Support Vector Machine. Via exhaustive searches, we found that the RF classifier with W2V consistently achieved better performance on different datasets. The two main controlling factors were: (i) the dataset-specific W2V dictionary was generated from the training and independent test datasets instead of the widely used general UniProt proteome and (ii) a systematic search was conducted and determined the optimal k-mer value in W2V, which provides greater discrimination between positive and negative samples. Therefore, our proposed method, named iACVP, consistently provides better prediction performance compared with existing state-of-the-art methods. To assist experimentalists in identifying putative ACVPs, we implemented our model as a web server accessible via the following link: http://kurata35.bio.kyutech.ac.jp/iACVP.

In molecular biology, it is a general assumption that the ensemble of expressed molecules, their activities and interactions determine biological function, cellular states and phenotypes. Stable protein complexes-or macromolecular machines-are, in turn, the key functional entities mediating and modulating most biological processes. Although identifying protein complexes and their subunit composition can now be done inexpensively and at scale, determining their function remains challenging and labor intensive. This study describes Protein Complex Function predictor (PCfun), the first computational framework for the systematic annotation of protein complex functions using Gene Ontology (GO) terms. PCfun is built upon a word embedding using natural language processing techniques based on 1 million open access PubMed Central articles. Specifically, PCfun leverages two approaches for accurately identifying protein complex function, including: (i) an unsupervised approach that obtains the nearest neighbor (NN) GO term word vectors for a protein complex query vector and (ii) a supervised approach using Random Forest (RF) models trained specifically for recovering the GO terms of protein complex queries described in the CORUM protein complex database. PCfun consolidates both approaches by performing a hypergeometric statistical test to enrich the top NN GO terms within the child terms of the GO terms predicted by the RF models. The documentation and implementation of the PCfun package are available at https://github.com/sharmavaruns/PCfun. We anticipate that PCfun will serve as a useful tool and novel paradigm for the large-scale characterization of protein complex function.

The identification of the conserved and variable regions in the multiple sequence alignment (MSA) is critical to accelerating the process of understanding the function of genes. MSA visualizations allow us to transform sequence features into understandable visual representations. As the sequence-structure-function relationship gains increasing attention in molecular biology studies, the simple display of nucleotide or protein sequence alignment is not satisfied. A more scalable visualization is required to broaden the scope of sequence investigation. Here we present ggmsa, an R package for mining comprehensive sequence features and integrating the associated data of MSA by a variety of display methods. To uncover sequence conservation patterns, variations and recombination at the site level, sequence bundles, sequence logos, stacked sequence alignment and comparative plots are implemented. ggmsa supports integrating the correlation of MSA sequences and their phenotypes, as well as other traits such as ancestral sequences, molecular structures, molecular functions and expression levels. We also design a new visualization method for genome alignments in multiple alignment format to explore the pattern of within and between species variation. Combining these visual representations with prime knowledge, ggmsa assists researchers in discovering MSA and making decisions. The ggmsa package is open-source software released under the Artistic-2.0 license, and it is freely available on Bioconductor (https://bioconductor.org/packages/ggmsa) and Github (https://github.com/YuLab-SMU/ggmsa).

Metabolism is the process by which an organism continuously replaces old substances with new substances. It plays an important role in maintaining human life, body growth and reproduction. More and more researchers have shown that the concentrations of some metabolites in patients are different from those in healthy people. Traditional biological experiments can test some hypotheses and verify their relationships but usually take a considerable amount of time and money. Therefore, it is urgent to develop a new computational method to identify the relationships between metabolites and diseases. In this work, we present a new deep learning algorithm named as graph convolutional network with graph attention network (GCNAT) to predict the potential associations of disease-related metabolites. First, we construct a heterogeneous network based on known metabolite-disease associations, metabolite-metabolite similarities and disease-disease similarities. Metabolite and disease features are encoded and learned through the graph convolutional neural network. Then, a graph attention layer is used to combine the embeddings of multiple convolutional layers, and the corresponding attention coefficients are calculated to assign different weights to the embeddings of each layer. Further, the prediction result is obtained by decoding and scoring the final synthetic embeddings. Finally, GCNAT achieves a reliable area under the receiver operating characteristic curve of 0.95 and the precision-recall curve of 0.405, which are better than the results of existing five state-of-the-art predictive methods in 5-fold cross-validation, and the case studies show that the metabolite-disease correlations predicted by our method can be successfully demonstrated by relevant experiments. We hope that GCNAT could be a useful biomedical research tool for predicting potential metabolite-disease associations in the future.

A representation method is an algorithm that calculates numerical feature vectors for samples in a dataset. Such vectors, also known as embeddings, define a relatively low-dimensional space able to efficiently encode high-dimensional data. Very recently, many types of learned data representations based on machine learning have appeared and are being applied to several tasks in bioinformatics. In particular, protein representation learning methods integrate different types of protein information (sequence, domains, etc.), in supervised or unsupervised learning approaches, and provide embeddings of protein sequences that can be used for downstream tasks. One task that is of special interest is the automatic function prediction of the huge number of novel proteins that are being discovered nowadays and are still totally uncharacterized. However, despite its importance, up to date there is not a fair benchmark study of the predictive performance of existing proposals on the same large set of proteins and for very concrete and common bioinformatics tasks. Therefore, this lack of benchmark studies prevent the community from using adequate predictive methods for accelerating the functional characterization of proteins. In this study, we performed a detailed comparison of protein sequence representation learning methods, explaining each approach and comparing them with an experimental benchmark on several bioinformatics tasks: (i) determining protein sequence similarity in the embedding space; (ii) inferring protein domains and (iii) predicting ontology-based protein functions. We examine the advantages and disadvantages of each representation approach over the benchmark results. We hope the results and the discussion of this study can help the community to select the most adequate machine learning-based technique for protein representation according to the bioinformatics task at hand.

In this study, we introduce an interpretable graph-based deep learning prediction model, AttentionSiteDTI, which utilizes protein binding sites along with a self-attention mechanism to address the problem of drug-target interaction prediction. Our proposed model is inspired by sentence classification models in the field of Natural Language Processing, where the drug-target complex is treated as a sentence with relational meaning between its biochemical entities a.k.a. protein pockets and drug molecule. AttentionSiteDTI enables interpretability by identifying the protein binding sites that contribute the most toward the drug-target interaction. Results on three benchmark datasets show improved performance compared with the current state-of-the-art models. More significantly, unlike previous studies, our model shows superior performance, when tested on new proteins (i.e. high generalizability). Through multidisciplinary collaboration, we further experimentally evaluate the practical potential of our proposed approach. To achieve this, we first computationally predict the binding interactions between some candidate compounds and a target protein, then experimentally validate the binding interactions for these pairs in the laboratory. The high agreement between the computationally predicted and experimentally observed (measured) drug-target interactions illustrates the potential of our method as an effective pre-screening tool in drug repurposing applications.

RNA binding proteins (RBPs) are critical for the post-transcriptional control of RNAs and play vital roles in a myriad of biological processes, such as RNA localization and gene regulation. Therefore, computational methods that are capable of accurately identifying RBPs are highly desirable and have important implications for biomedical and biotechnological applications. Here, we propose a two-stage deep transfer learning-based framework, termed RBP-TSTL, for accurate prediction of RBPs. In the first stage, the knowledge from the self-supervised pre-trained model was extracted as feature embeddings and used to represent the protein sequences, while in the second stage, a customized deep learning model was initialized based on an annotated pre-training RBPs dataset before being fine-tuned on each corresponding target species dataset. This two-stage transfer learning framework can enable the RBP-TSTL model to be effectively trained to learn and improve the prediction performance. Extensive performance benchmarking of the RBP-TSTL models trained using the features generated by the self-supervised pre-trained model and other models trained using hand-crafting encoding features demonstrated the effectiveness of the proposed two-stage knowledge transfer strategy based on the self-supervised pre-trained models. Using the best-performing RBP-TSTL models, we further conducted genome-scale RBP predictions for Homo sapiens, Arabidopsis thaliana, Escherichia coli, and Salmonella and established a computational compendium containing all the predicted putative RBPs candidates. We anticipate that the proposed RBP-TSTL approach will be explored as a useful tool for the characterization of RNA-binding proteins and exploration of their sequence-structure-function relationships.

Recent technological advances have led to an exponential expansion of biological sequence data and extraction of meaningful information through Machine Learning (ML) algorithms. This knowledge has improved the understanding of mechanisms related to several fatal diseases, e.g. Cancer and coronavirus disease 2019, helping to develop innovative solutions, such as CRISPR-based gene editing, coronavirus vaccine and precision medicine. These advances benefit our society and economy, directly impacting people's lives in various areas, such as health care, drug discovery, forensic analysis and food processing. Nevertheless, ML-based approaches to biological data require representative, quantitative and informative features. Many ML algorithms can handle only numerical data, and therefore sequences need to be translated into a numerical feature vector. This process, known as feature extraction, is a fundamental step for developing high-quality ML-based models in bioinformatics, by allowing the feature engineering stage, with design and selection of suitable features. Feature engineering, ML algorithm selection and hyperparameter tuning are often manual and time-consuming processes, requiring extensive domain knowledge. To deal with this problem, we present a new package: BioAutoML. BioAutoML automatically runs an end-to-end ML pipeline, extracting numerical and informative features from biological sequence databases, using the MathFeature package, and automating the feature selection, ML algorithm(s) recommendation and tuning of the selected algorithm(s) hyperparameters, using Automated ML (AutoML). BioAutoML has two components, divided into four modules: (1) automated feature engineering (feature extraction and selection modules) and (2) Metalearning (algorithm recommendation and hyper-parameter tuning modules). We experimentally evaluate BioAutoML in two different scenarios: (i) prediction of the three main classes of noncoding RNAs (ncRNAs) and (ii) prediction of the eight categories of ncRNAs in bacteria, including housekeeping and regulatory types. To assess BioAutoML predictive performance, it is experimentally compared with two other AutoML tools (RECIPE and TPOT). According to the experimental results, BioAutoML can accelerate new studies, reducing the cost of feature engineering processing and either keeping or improving predictive performance. BioAutoML is freely available at https://github.com/Bonidia/BioAutoML.

Drug target discovery is an essential step to reveal the mechanism of action (MoA) underlying drug therapeutic effects and/or side effects. Most of the approaches are usually labor-intensive while unable to identify the tissue-specific interacting targets, especially the targets with weaker drug binding affinity. In this work, we proposed an integrated pipeline, FL-DTD, to predict the drug interacting targets of novel compounds in a tissue-specific manner. This method was built based on a hypothesis that cells under a status of homeostasis would take responses to drug perturbation by activating feedback loops. Therefore, the drug interacting targets can be predicted by analyzing the network responses after drug perturbation. We evaluated this method using the expression data of estrogen stimulation, gene manipulation and drug perturbation and validated its good performance to identify the annotated drug targets. Using STAT3 as a target protein, we applied this method to drug perturbation data of 500 natural compounds and predicted five compounds with STAT3 interacting activities. Experimental assay validated the STAT3-interacting activities of four compounds. Overall, our evaluation suggests that FL-DTD predicts the drug interacting targets with good accuracy and can be used for drug target discovery.

As the development of new drugs reaches its physical and financial limits, drug repurposing has become more important than ever. For mechanistically grounded drug repurposing, it is crucial to uncover the disease mechanisms and to detect clusters of mechanistically related diseases. Various methods for computing candidate disease mechanisms and disease clusters exist. However, in the absence of ground truth, in silico validation is challenging. This constitutes a major hurdle toward the adoption of in silico prediction tools by experimentalists who are often hesitant to carry out wet-lab validations for predicted candidate mechanisms without clearly quantified initial plausibility. To address this problem, we present DIGEST (in silico validation of disease and gene sets, clusterings or subnetworks), a Python-based validation tool available as a web interface (https://digest-validation.net), as a stand-alone package or over a REST API. DIGEST greatly facilitates in silico validation of gene and disease sets, clusterings or subnetworks via fully automated pipelines comprising disease and gene ID mapping, enrichment analysis, comparisons of shared genes and variants and background distribution estimation. Moreover, functionality is provided to automatically update the external databases used by the pipelines. DIGEST hence allows the user to assess the statistical significance of candidate mechanisms with regard to functional and genetic coherence and enables the computation of empirical $P$-values with just a few mouse clicks.

Unintended inhibition of the human ether-à-go-go-related gene (hERG) ion channel by small molecules leads to severe cardiotoxicity. Thus, hERG channel blockage is a significant concern in the development of new drugs. Several computational models have been developed to predict hERG channel blockage, including deep learning models; however, they lack robustness, reliability and interpretability. Here, we developed a graph-based Bayesian deep learning model for hERG channel blocker prediction, named BayeshERG, which has robust predictive power, high reliability and high resolution of interpretability. First, we applied transfer learning with 300 000 large data in initial pre-training to increase the predictive performance. Second, we implemented a Bayesian neural network with Monte Carlo dropout to calibrate the uncertainty of the prediction. Third, we utilized global multihead attentive pooling to augment the high resolution of structural interpretability for the hERG channel blockers and nonblockers. We conducted both internal and external validations for stringent evaluation; in particular, we benchmarked most of the publicly available hERG channel blocker prediction models. We showed that our proposed model outperformed predictive performance and uncertainty calibration performance. Furthermore, we found that our model learned to focus on the essential substructures of hERG channel blockers via an attention mechanism. Finally, we validated the prediction results of our model by conducting in vitro experiments and confirmed its high validity. In summary, BayeshERG could serve as a versatile tool for discovering hERG channel blockers and helping maximize the possibility of successful drug discovery. The data and source code are available at our GitHub repository (https://github.com/GIST-CSBL/BayeshERG).

Protein phylogenetic analysis focuses on the evolutionary relationships among related protein sequences and can help researchers infer protein functions and developmental trajectories. With the advent of the big data era, the existing protein phylogenetic methods, including distance matrix and character-based methods, are facing challenges in both running time and application scope. Here, we developed an R package that we call CProtMEDIAS that is useful for protein phylogenetic analysis. In contrast to existing phylogenetic analysis methods, CProtMEDIAS utilizes dimensionality reduction algorithms to digitize multiple sequence alignments and quickly conduct phylogenetic analysis with a large number of amino acid sequences from similarly distant protein families and species. We used CProtMEDIAS to perform a dimensionality reduction, clustering, pseudotime, specific residue and evolutionary trajectory analysis of the plant homeobox superfamily. We found that CProtMEDIAS delivers consistent clustering, fast running and elegant presentation and thus provides powerful new tools and methods for protein clustering and evolutionary analysis.

BACKGROUND: Angiogenesis is regulated by multiple genes whose variants can lead to different disorders. Among them, rare diseases are a heterogeneous group of pathologies, most of them genetic, whose information may be of interest to determine the still unknown genetic and molecular causes of other diseases. In this work, we use the information on rare diseases dependent on angiogenesis to investigate the genes that are associated with this biological process and to determine if there are interactions between the genes involved in its deregulation. RESULTS: We propose a systemic approach supported by the use of pathological phenotypes to group diseases by semantic similarity. We grouped 158 angiogenesis-related rare diseases in 18 clusters based on their phenotypes. Of them, 16 clusters had traceable gene connections in a high-quality interaction network. These disease clusters are associated with 130 different genes. We searched for genes associated with angiogenesis througth ClinVar pathogenic variants. Of the seven retrieved genes, our system confirms six of them. Furthermore, it allowed us to identify common affected functions among these disease clusters. AVAILABILITY: https://github.com/ElenaRojano/angio_cluster. CONTACT: seoanezonjic@uma.es and elenarojano@uma.es.

Accurate prediction of open reading frames (ORFs) is important for studying and using genome sequences. Ribosomes move along mRNA strands with a step of three nucleotides and datasets carrying this information can be used to predict ORFs. The ribosome-protected footprints (RPFs) feature a significant 3-nt periodicity on mRNAs and are powerful in predicting translating ORFs, including small ORFs (sORFs), but the application of RPFs is limited because they are too short to be accurately mapped in complex genomes. In this study, we found a significant 3-nt periodicity in the datasets of populational genomic variants in coding sequences, in which the nucleotide diversity increases every three nucleotides. We suggest that this feature can be used to predict ORFs and develop the Python package 'OrfPP', which recovers ~83% of the annotated ORFs in the tested genomes on average, independent of the population sizes and the complexity of the genomes. The novel ORFs, including sORFs, identified from single-nucleotide polymorphisms are supported by protein mass spectrometry evidence comparable to that of the annotated ORFs. The application of OrfPP to tetraploid cotton and hexaploid wheat genomes successfully identified 76.17% and 87.43% of the annotated ORFs in the genomes, respectively, as well as 4704 sORFs, including 1182 upstream and 2110 downstream ORFs in cotton and 5025 sORFs, including 232 upstream and 234 downstream ORFs in wheat. Overall, we propose an alternative and supplementary approach for ORF prediction that can extend the studies of sORFs to more complex genomes.

Carcinomas are complex ecosystems composed of cancer, stromal and immune cells. Communication between these cells and their microenvironments induces cancer progression and causes therapy resistance. In order to improve the treatment of cancers, it is essential to quantify crosstalk between and within various cell types in a tumour microenvironment. Focusing on the coordinated expression patterns of ligands and cognate receptors, cell-cell communication can be inferred through ligand-receptor interactions (LRIs). In this manuscript, we carry out the following work: (i) introduce pipeline for ligand-receptor-mediated intercellular communication estimation from single-cell transcriptomics and list a few available LRI-related databases and visualization tools; (ii) demonstrate seven classical intercellular communication scoring strategies, highlight four types of representative intercellular communication inference methods, including network-based approaches, machine learning-based approaches, spatial information-based approaches and other approaches; (iii) summarize the evaluation and validation avenues for intercellular communication inference and analyze the advantages and limitations for the above four types of cell-cell communication methods; (iv) comment several major challenges while provide further research directions for intercellular communication analysis in the tumour microenvironments. We anticipate that this work helps to better understand intercellular crosstalk and to further develop powerful cell-cell communication estimation tools for tumor-targeted therapy.

With the advances in high-throughput biotechnologies, high-dimensional multi-layer omics data become increasingly available. They can provide both confirmatory and complementary information to disease risk and thus have offered unprecedented opportunities for risk prediction studies. However, the high-dimensionality and complex inter/intra-relationships among multi-omics data have brought tremendous analytical challenges. Here we present a computationally efficient penalized linear mixed model with generalized method of moments estimator (MpLMMGMM) for the prediction analysis on multi-omics data. Our method extends the widely used linear mixed model proposed for genomic risk predictions to model multi-omics data, where kernel functions are used to capture various types of predictive effects from different layers of omics data and penalty terms are introduced to reduce the impact of noise. Compared with existing penalized linear mixed models, the proposed method adopts the generalized method of moments estimator and it is much more computationally efficient. Through extensive simulation studies and the analysis of positron emission tomography imaging outcomes, we have demonstrated that MpLMMGMM can simultaneously consider a large number of variables and efficiently select those that are predictive from the corresponding omics layers. It can capture both linear and nonlinear predictive effects and achieves better prediction performance than competing methods.

In the previous study, we developed the generalized drug-induced liver injury (DILI) prediction model-ResNet18DNN to predict DILI based on multi-source combined DILI dataset and achieved better performance than that of previously published described DILI prediction models. Recently, we were honored to receive the invitation from the editor to response the Letter to Editor by Liu Zhichao, et al. We were glad that our research has attracted the attention of Liu's team and they has put forward their opinions on our research. In this response to Letter to the Editor, we will respond to these comments.

Protein kinases play crucial roles in many cellular signaling processes, making them become important targets for drug discovery. But drug resistance mediated by mutation puts a barrier to the therapeutic effect of kinase inhibitors. Fragment-based drug discovery has been successfully applied to overcome such resistance. However, the complicate kinase-inhibitor fragment interaction and fragment-to-lead process seriously limit the efficiency of kinase inhibitor discovery against resistance caused by mutation. Here, we constructed a comprehensive web platform KinaFrag for the fragment-based kinase inhibitor discovery to overcome resistance. The kinase-inhibitor fragment space was investigated from 7783 crystal kinase-inhibitor fragment complexes, and the structural requirements of kinase subpockets were analyzed. The core fragment-based virtual screening workflow towards specific subpockets was developed to generate new kinase inhibitors. A series of tropomyosin receptor kinase (TRK) inhibitors were designed, and the most potent compound YT9 exhibits up to 70-fold activity improvement than marketed drugs larotrectinib and selitrectinib against G595R, G667C and F589L mutations of TRKA. YT9 shows promising antiproliferative against tumor cells in vitro and effectively inhibits tumor growth in vivo for wild type TRK and TRK mutants. Our results illustrate the great potential of KinaFrag in the kinase inhibitor discovery to combat resistance mediated by mutation. KinaFrag is freely available at http://chemyang.ccnu.edu.cn/ccb/database/KinaFrag/.

Abnormal accumulation of R-loops results in replication stress, genome instability, chromatin alterations and gene silencing. Little research has been done to characterize functional relationships among R-loops, histone marks, RNA polymerase II (RNAPII) transcription and gene regulation. We built extremely randomized trees (ETs) models to predict the genome-wide R-loops using RNAPII and multiple histone modifications chromatin immunoprecipitation (ChIP)-seq, DNase-seq, Global Run-On sequencing (GRO-seq) and R-loop profiling data. We compared the performance of ET models to multiple machine learning approaches, and the proposed ET models achieved the best and extremely robust performances. Epigenetic profiles are highly predictive of R-loops genome-widely and they are strongly associated with R-loop formation. In addition, the presence of R-loops is significantly correlated with RNAPII transcription activity, H3K4me3 and open chromatin around the transcription start site, and H3K9me1 and H3K9me3 around the transcription termination site. RNAPII pausing defects were correlated with 5'R-loops accumulation, and transcriptional termination defects and read-throughs were correlated with 3'R-loops accumulation. Furthermore, we found driver genes with 5'R-loops and RNAPII pausing defects express significantly higher and genes with 3'R-loops and read-through transcription express significantly lower than genes without R-loops. These driver genes are enriched with chromosomal instability, Hippo-Merlin signaling Dysregulation, DNA damage response and TGF-β pathways, indicating R-loops accumulating at the 5' end of genes play oncogenic roles, whereas at the 3' end of genes play tumor-suppressive roles in tumorigenesis.

Exploring multimorbidity relationships among diseases is of great importance for understanding their shared mechanisms, precise diagnosis and treatment. However, the landscape of multimorbidities is still far from complete due to the complex nature of multimorbidity. Although various types of biological data, such as biomolecules and clinical symptoms, have been used to identify multimorbidities, the population phenotype information (e.g. physical activity and diet) remains less explored for multimorbidity. Here, we present a graph convolutional network (GCN) model, named MorbidGCN, for multimorbidity prediction by integrating population phenotypes and disease network. Specifically, MorbidGCN treats the multimorbidity prediction as a missing link prediction problem in the disease network, where a novel feature selection method is embedded to select important phenotypes. Benchmarking results on two large-scale multimorbidity data sets, i.e. the UK Biobank (UKB) and Human Disease Network (HuDiNe) data sets, demonstrate that MorbidGCN outperforms other competitive methods. With MorbidGCN, 9742 and 14 010 novel multimorbidities are identified in the UKB and HuDiNe data sets, respectively. Moreover, we notice that the selected phenotypes that are generally differentially distributed between multimorbidity patients and single-disease patients can help interpret multimorbidities and show potential for prognosis of multimorbidities.

AlphaFold2 can predict protein complex structures as long as a multiple sequence alignment (MSA) of the interologs of the target protein-protein interaction (PPI) can be provided. In this study, a simplified phylogeny-based approach was applied to generate the MSA of interologs, which was then used as the input to AlphaFold2 for protein complex structure prediction. In this extensively benchmarked protocol on nonredundant PPI dataset, including 107 bacterial PPIs and 442 eukaryotic PPIs, we show complex structures of 79.5% of the bacterial PPIs and 49.8% of the eukaryotic PPIs can be successfully predicted, which yielded significantly better performance than the application of MSA of interologs prepared by two existing approaches. Considering PPIs may not be conserved in species with long evolutionary distances, we further restricted interologs in the MSA to different taxonomic ranks of the species of the target PPI in protein complex structure prediction. We found that the success rates can be increased to 87.9% for the bacterial PPIs and 56.3% for the eukaryotic PPIs if interologs in the MSA are restricted to a specific taxonomic rank of the species of each target PPI. Finally, we show that the optimal taxonomic ranks for protein complex structure prediction can be selected with the application of the predicted template modeling (TM) scores of the output models.

Deep learning is an artificial intelligence technique in which models express geometric transformations over multiple levels. This method has shown great promise in various fields, including drug development. The availability of public structure databases prompted the researchers to use generative artificial intelligence models to narrow down their search of the chemical space, a novel approach to chemogenomics and de novo drug development. In this study, we developed a strategy that combined an accelerated LSTM_Chem (long short-term memory for de novo compounds generation), dense fully convolutional neural network (DFCNN), and docking to generate a large number of de novo small molecular chemical compounds for given targets. To demonstrate its efficacy and applicability, six important targets that account for various human disorders were used as test examples. Moreover, using the M protease as a proof-of-concept example, we find that iteratively training with previously selected candidates can significantly increase the chance of obtaining novel compounds with higher and higher predicted binding affinities. In addition, we also check the potential benefit of obtaining reliable final de novo compounds with the help of MD simulation and metadynamics simulation. The generation of de novo compounds and the discovery of binders against various targets proposed here would be a practical and effective approach. Assessing the efficacy of these top de novo compounds with biochemical studies is promising to promote related drug development.

Proteins maintain the functional order of cell in life by interacting with other proteins. Determination of protein complex structural information gives biological insights for the research of diseases and drugs. Recently, a breakthrough has been made in protein monomer structure prediction. However, due to the limited number of the known protein structure and homologous sequences of complexes, the prediction of residue-residue contacts on hetero-dimer interfaces is still a challenge. In this study, we have developed a deep learning framework for inferring inter-protein residue contacts from sequential information, called HDIContact. We utilized transfer learning strategy to produce Multiple Sequence Alignment (MSA) two-dimensional (2D) embedding based on patterns of concatenated MSA, which could reduce the influence of noise on MSA caused by mismatched sequences or less homology. For MSA 2D embedding, HDIContact took advantage of Bi-directional Long Short-Term Memory (BiLSTM) with two-channel to capture 2D context of residue pairs. Our comprehensive assessment on the Escherichia coli (E. coli) test dataset showed that HDIContact outperformed other state-of-the-art methods, with top precision of 65.96%, the Area Under the Receiver Operating Characteristic curve (AUROC) of 83.08% and the Area Under the Precision Recall curve (AUPR) of 25.02%. In addition, we analyzed the potential of HDIContact for human-virus protein-protein complexes, by achieving top five precision of 80% on O75475-P04584 related to Human Immunodeficiency Virus. All experiments indicated that our method was a valuable technical tool for predicting inter-protein residue contacts, which would be helpful for understanding protein-protein interaction mechanisms.

The pan-genome analysis of bacteria provides detailed insight into the diversity and evolution of a bacterial population. However, the genomes involved in the pan-genome analysis should be checked carefully, as the inclusion of confounding strains would have unfavorable effects on the identification of core genes, and the highly similar strains could bias the results of the pan-genome state (open versus closed). In this study, we found that the inclusion of highly similar strains also affects the results of unique genes in pan-genome analysis, which leads to a significant underestimation of the number of unique genes in the pan-genome. Therefore, these strains should be excluded from pan-genome analysis at the early stage of data processing. Currently, tens of thousands of genomes have been sequenced for Escherichia coli, which provides an unprecedented opportunity as well as a challenge for pan-genome analysis of this classical model organism. Using the proposed strategies, a high-quality E. coli pan-genome was obtained, and the unique genes was extracted and analyzed, revealing an association between the unique gene clusters and genomic islands from a pan-genome perspective, which may facilitate the identification of genomic islands.

As the most abundant RNA modification, N6-methyladenosine (m6A) plays an important role in various RNA activities including gene expression and translation. With the rapid application of MeRIP-seq technology, samples of multiple groups, such as the involved multiple viral/ bacterial infection or distinct cell differentiation stages, are extracted from same experimental unit. However, our current knowledge about how the dynamic m6A regulating gene expression and the role in certain biological processes (e.g. immune response in this complex context) is largely elusive due to lack of effective tools. To address this issue, we proposed a Bayesian hierarchical mixture model (called m6Aexpress-BHM) to predict m6A regulation of gene expression (m6A-reg-exp) in multiple groups of MeRIP-seq experiment with limited samples. Comprehensive evaluations of m6Aexpress-BHM on the simulated data demonstrate its high predicting precision and robustness. Applying m6Aexpress-BHM on three real-world datasets (i.e. Flaviviridae infection, infected time-points of bacteria and differentiation stages of dendritic cells), we predicted more m6A-reg-exp genes with positive regulatory mode that significantly participate in innate immune or adaptive immune pathways, revealing the underlying mechanism of the regulatory function of m6A during immune response. In addition, we also found that m6A may influence the expression of PD-1/PD-L1 via regulating its interacted genes. These results demonstrate the power of m6Aexpress-BHM, helping us understand the m6A regulatory function in immune system.

Ribosomal deoxyribonucleic acid (DNA) (rDNA) repeats are tandemly located on five acrocentric chromosomes with up to hundreds of copies in the human genome. DNA methylation, the most well-studied epigenetic mechanism, has been characterized for most genomic regions across various biological contexts. However, rDNA methylation patterns remain largely unexplored due to the repetitive structure. In this study, we designed a specific mapping strategy to investigate rDNA methylation patterns at each CpG site across various physiological and pathological processes. We found that CpG sites on rDNA could be categorized into two types. One is within or adjacent to transcribed regions; the other is distal to transcribed regions. The former shows highly variable methylation levels across samples, while the latter shows stable high methylation levels in normal tissues but severe hypomethylation in tumors. We further showed that rDNA methylation profiles in plasma cell-free DNA could be used as a biomarker for cancer detection. It shows good performances on public datasets, including colorectal cancer [area under the curve (AUC) = 0.85], lung cancer (AUC = 0.84), hepatocellular carcinoma (AUC = 0.91) and in-house generated hepatocellular carcinoma dataset (AUC = 0.96) even at low genome coverage (<1×). Taken together, these findings broaden our understanding of rDNA regulation and suggest the potential utility of rDNA methylation features as disease biomarkers.

Cell-free DNA (cfDNA) provides a convenient diagnosis avenue for noninvasive cancer detection. The current methods are focused on identifying circulating tumor DNA (ctDNA)s genomic aberrations, e.g. mutations, copy number aberrations (CNAs) or methylation changes. In this study, we report a new computational method that unifies two orthogonal pieces of information, namely methylation and CNAs, derived from whole-genome bisulfite sequencing (WGBS) data to quantify low tumor content in cfDNA. It implements a Bayes model to enrich ctDNA from WGBS data based on hypomethylation haplotypes, and subsequently, models CNAs for cancer detection. We generated WGBS data in a total of 262 samples, including high-depth (>20×, deduped high mapping quality reads) data in 76 samples with matched triplets (tumor, adjacent normal and cfDNA) and low-depth (~2.5×, deduped high mapping quality reads) data in 186 samples. We identified a total of 54 Mb regions of hypomethylation haplotypes for model building, a vast majority of which are not covered in the HumanMethylation450 arrays. We showed that our model is able to substantially enrich ctDNA reads (tens of folds), with clearly elevated CNAs that faithfully match the CNAs in the paired tumor samples. In the 19 hepatocellular carcinoma cfDNA samples, the estimated enrichment is as high as 16 fold, and in the simulation data, it can achieve over 30-fold enrichment for a ctDNA level of 0.5% with a sequencing depth of 600×. We also found that these hypomethylation regions are also shared among many cancer types, thus demonstrating the potential of our framework for pancancer early detection.

Rational vaccine design, especially vaccine antigen identification and optimization, is critical to successful and efficient vaccine development against various infectious diseases including coronavirus disease 2019 (COVID-19). In general, computational vaccine design includes three major stages: (i) identification and annotation of experimentally verified gold standard protective antigens through literature mining, (ii) rational vaccine design using reverse vaccinology (RV) and structural vaccinology (SV) and (iii) post-licensure vaccine success and adverse event surveillance and its usage for vaccine design. Protegen is a database of experimentally verified protective antigens, which can be used as gold standard data for rational vaccine design. RV predicts protective antigen targets primarily from genome sequence analysis. SV refines antigens through structural engineering. Recently, RV and SV approaches, with the support of various machine learning methods, have been applied to COVID-19 vaccine design. The analysis of post-licensure vaccine adverse event report data also provides valuable results in terms of vaccine safety and how vaccines should be used or paused. Ontology standardizes and incorporates heterogeneous data and knowledge in a human- and computer-interpretable manner, further supporting machine learning and vaccine design. Future directions on rational vaccine design are discussed.

Geographic patterns of human genetic variation provide important insights into human evolution and disease. A commonly used tool to detect and describe them is principal component analysis (PCA) or the supervised linear discriminant analysis of principal components (DAPC). However, genetic features produced from both approaches could fail to correctly characterize population structure for complex scenarios involving admixture. In this study, we introduce Kernel Local Fisher Discriminant Analysis of Principal Components (KLFDAPC), a supervised non-linear approach for inferring individual geographic genetic structure that could rectify the limitations of these approaches by preserving the multimodal space of samples. We tested the power of KLFDAPC to infer population structure and to predict individual geographic origin using neural networks. Simulation results showed that KLFDAPC has higher discriminatory power than PCA and DAPC. The application of our method to empirical European and East Asian genome-wide genetic datasets indicated that the first two reduced features of KLFDAPC correctly recapitulated the geography of individuals and significantly improved the accuracy of predicting individual geographic origin when compared to PCA and DAPC. Therefore, KLFDAPC can be useful for geographic ancestry inference, design of genome scans and correction for spatial stratification in GWAS that link genes to adaptation or disease susceptibility.

Predicting drug-target interactions (DTIs) is crucial at many phases of drug discovery and repositioning. Many computational methods based on heterogeneous networks (HNs) have proved their potential to predict DTIs by capturing extensive biological knowledge and semantic information from meta-paths. However, existing methods manually customize meta-paths, which is overly dependent on some specific expertise. Such strategy heavily limits the scalability and flexibility of these models, and even affects their predictive performance. To alleviate this limitation, we propose a novel HN-based method with attentive meta-path extraction for DTI prediction, named HampDTI, which is capable of automatically extracting useful meta-paths through a learnable attention mechanism instead of pre-definition based on domain knowledge. Specifically, by scoring multi-hop connections across various relations in the HN with each relation assigned an attention weight, HampDTI constructs a new trainable graph structure, called meta-path graph. Such meta-path graph implicitly measures the importance of every possible meta-path between drugs and targets. To enable HampDTI to extract more diverse meta-paths, we adopt a multi-channel mechanism to generate multiple meta-path graphs. Then, a graph neural network is deployed on the generated meta-path graphs to yield the multi-channel embeddings of drugs and targets. Finally, HampDTI fuses all embeddings from different channels for predicting DTIs. The meta-path graphs are optimized along with the model training such that HampDTI can adaptively extract valuable meta-paths for DTI prediction. The experiments on benchmark datasets not only show the superiority of HampDTI in DTI prediction over several baseline methods, but also, more importantly, demonstrate the effectiveness of the model discovering important meta-paths.

The three-dimensional (3D) chromosomal structure plays an essential role in all DNA-templated processes, including gene transcription, DNA replication and other cellular processes. Although developing chromosome conformation capture (3C) methods, such as Hi-C, which can generate chromosomal contact data characterized genome-wide chromosomal structural properties, understanding 3D genomic nature-based on Hi-C data remains lacking. Here, we propose a persistent spectral simplicial complex (PerSpectSC) model to describe Hi-C data for the first time. Specifically, a filtration process is introduced to generate a series of nested simplicial complexes at different scales. For each of these simplicial complexes, its spectral information can be calculated from the corresponding Hodge Laplacian matrix. PerSpectSC model describes the persistence and variation of the spectral information of the nested simplicial complexes during the filtration process. Different from all previous models, our PerSpectSC-based features provide a quantitative global-scale characterization of chromosome structures and topology. Our descriptors can successfully classify cell types and also cellular differentiation stages for all the 24 types of chromosomes simultaneously. In particular, persistent minimum best characterizes cell types and Dim (1) persistent multiplicity best characterizes cellular differentiation. These results demonstrate the great potential of our PerSpectSC-based models in polymeric data analysis.

Many studies have proved that small nucleolar RNAs (snoRNAs) play critical roles in the development of various human complex diseases. Discovering the associations between snoRNAs and diseases is an important step toward understanding the pathogenesis and characteristics of diseases. However, uncovering associations via traditional experimental approaches is costly and time-consuming. This study proposed a bounded nuclear norm regularization-based method, called PSnoD, to predict snoRNA-disease associations. Benchmark experiments showed that compared with the state-of-the-art methods, PSnoD achieved a superior performance in the 5-fold stratified shuffle split. PSnoD produced a robust performance with an area under receiver-operating characteristic of 0.90 and an area under precision-recall of 0.55, highlighting the effectiveness of our proposed method. In addition, the computational efficiency of PSnoD was also demonstrated by comparison with other matrix completion techniques. More importantly, the case study further elucidated the ability of PSnoD to screen potential snoRNA-disease associations. The code of PSnoD has been uploaded to https://github.com/linDing-groups/PSnoD. Based on PSnoD, we established a web server that is freely accessed via http://psnod.lin-group.cn/.

Spatial transcriptomics (ST) technologies allow researchers to examine transcriptional profiles along with maintained positional information. Such spatially resolved transcriptional characterization of intact tissue samples provides an integrated view of gene expression in its natural spatial and functional context. However, high-throughput sequencing-based ST technologies cannot yet reach single cell resolution. Thus, similar to bulk RNA-seq data, gene expression data at ST spot-level reflect transcriptional profiles of multiple cells and entail the inference of cell-type composition within each ST spot for valid and powerful subsequent analyses. Realizing the critical importance of cell-type decomposition, multiple groups have developed ST deconvolution methods. The aim of this work is to review state-of-the-art methods for ST deconvolution, comparing their strengths and weaknesses. In particular, we construct ST spots from single-cell level ST data to assess the performance of 10 methods, with either ideal reference or non-ideal reference. Furthermore, we examine the performance of these methods on spot- and bead-level ST data by comparing estimated cell-type proportions to carefully matched single-cell ST data. In comparing the performance on various tissues and technological platforms, we concluded that RCTD and stereoscope achieve more robust and accurate inferences.

Long noncoding RNAs (lncRNAs) are primarily regulated by their cellular localization, which is responsible for their molecular functions, including cell cycle regulation and genome rearrangements. Accurately identifying the subcellular location of lncRNAs from sequence information is crucial for a better understanding of their biological functions and mechanisms. In contrast to traditional experimental methods, bioinformatics or computational methods can be applied for the annotation of lncRNA subcellular locations in humans more effectively. In the past, several machine learning-based methods have been developed to identify lncRNA subcellular localization, but relevant work for identifying cell-specific localization of human lncRNA remains limited. In this study, we present the first application of the tree-based stacking approach, TACOS, which allows users to identify the subcellular localization of human lncRNA in 10 different cell types. Specifically, we conducted comprehensive evaluations of six tree-based classifiers with 10 different feature descriptors, using a newly constructed balanced training dataset for each cell type. Subsequently, the strengths of the AdaBoost baseline models were integrated via a stacking approach, with an appropriate tree-based classifier for the final prediction. TACOS displayed consistent performance in both the cross-validation and independent assessments compared with the other two approaches employed in this study. The user-friendly online TACOS web server can be accessed at https://balalab-skku.org/TACOS.

The evolution of drug-resistant pathogenic microbial species is a major global health concern. Naturally occurring, antimicrobial peptides (AMPs) are considered promising candidates to address antibiotic resistance problems. A variety of computational methods have been developed to accurately predict AMPs. The majority of such methods are not microbial strain specific (MSS): they can predict whether a given peptide is active against some microbe, but cannot accurately calculate whether such peptide would be active against a particular MS. Due to insufficient data on most MS, only a few MSS predictive models have been developed so far. To overcome this problem, we developed a novel approach that allows to improve MSS predictive models (MSSPM), based on properties, computed for AMP sequences and characteristics of genomes, computed for target MS. New models can perform predictions of AMPs for MS that do not have data on peptides tested on them. We tested various types of feature engineering as well as different machine learning (ML) algorithms to compare the predictive abilities of resulting models. Among the ML algorithms, Random Forest and AdaBoost performed best. By using genome characteristics as additional features, the performance for all models increased relative to models relying on AMP sequence-based properties only. Our novel MSS AMP predictor is freely accessible as part of DBAASP database resource at http://dbaasp.org/prediction/genome.

Mutations in human proteins lead to diseases. The structure of these proteins can help understand the mechanism of such diseases and develop therapeutics against them. With improved deep learning techniques, such as RoseTTAFold and AlphaFold, we can predict the structure of proteins even in the absence of structural homologs. We modeled and extracted the domains from 553 disease-associated human proteins without known protein structures or close homologs in the Protein Databank. We noticed that the model quality was higher and the Root mean square deviation (RMSD) lower between AlphaFold and RoseTTAFold models for domains that could be assigned to CATH families as compared to those which could only be assigned to Pfam families of unknown structure or could not be assigned to either. We predicted ligand-binding sites, protein-protein interfaces and conserved residues in these predicted structures. We then explored whether the disease-associated missense mutations were in the proximity of these predicted functional sites, whether they destabilized the protein structure based on ddG calculations or whether they were predicted to be pathogenic. We could explain 80% of these disease-associated mutations based on proximity to functional sites, structural destabilization or pathogenicity. When compared to polymorphisms, a larger percentage of disease-associated missense mutations were buried, closer to predicted functional sites, predicted as destabilizing and pathogenic. Usage of models from the two state-of-the-art techniques provide better confidence in our predictions, and we explain 93 additional mutations based on RoseTTAFold models which could not be explained based solely on AlphaFold models.