Tumor mutational burden (TMB) is a significant predictive biomarker for selecting patients that may benefit from immune checkpoint inhibitor therapy. Whole exome sequencing is a common method for measuring TMB; however, its clinical application is limited by the high cost and time-consuming wet-laboratory experiments and bioinformatics analysis. To address this challenge, we downloaded multimodal data of 326 gastric cancer patients from The Cancer Genome Atlas, including histopathological images, clinical data and various molecular data. Using these data, we conducted a comprehensive analysis to investigate the relationship between TMB, clinical factors, gene expression and image features extracted from hematoxylin and eosin images. We further explored the feasibility of predicting TMB levels, i.e. high and low TMB, by utilizing a residual network (Resnet)-based deep learning algorithm for histopathological image analysis. Moreover, we developed a multimodal fusion deep learning model that combines histopathological images with omics data to predict TMB levels. We evaluated the performance of our models against various state-of-the-art methods using different TMB thresholds and obtained promising results. Specifically, our histopathological image analysis model achieved an area under curve (AUC) of 0.749. Notably, the multimodal fusion model significantly outperformed the model that relied only on histopathological images, with the highest AUC of 0.971. Our findings suggest that histopathological images could be used with reasonable accuracy to predict TMB levels in gastric cancer patients, while multimodal deep learning could achieve even higher levels of accuracy. This study sheds new light on predicting TMB in gastric cancer patients.

Esophageal cancer (ESCA) has a bad prognosis. Long non-coding RNA (lncRNA) impacts on cell proliferation. However, the prognosis function of N6-methyladenosine (m6A)-associated lncRNAs (m6A-lncRNAs) in ESCA remains unknown. Univariate Cox analysis was applied to investigate prognosis related m6A-lncRNAs, based on which the samples were clustered. Wilcoxon rank and Chi-square tests were adopted to compare the clinical traits, survival, pathway activity and immune infiltration in different clusters where overall survival, clinical traits (N stage), tumor-invasive immune cells and pathway activity were found significantly different. Through least absolute shrinkage and selection operator and proportional hazard (Lasso-Cox) model, five m6A-lncRNAs were selected to construct the prognostic signature (m6A-lncSig) and risk score. To investigate the link between risk score and clinical traits or immunological microenvironments, Chi-square test and Spearman correlation analysis were utilized. Risk score was found connected with N stage, tumor stage, different clusters, macrophages M2, B cells naive and T cells CD4 memory resting. Risk score and tumor stage were found as independent prognostic variables. And the constructed nomogram model had high accuracy in predicting prognosis. The obtained m6A-lncSig could be taken as potential prognostic biomarker for ESCA patients. This study offers a theoretical foundation for clinical diagnosis and prognosis of ESCA.

In recent years, the emergence and development of single-cell sequencing technologies have provided unprecedented opportunities to analyze deoxyribonucleic acid, ribonucleic acid and proteins at single-cell resolution. The advancements and reduced costs of high-throughput technologies allow for parallel sequencing of multiple molecular layers from a single cell, providing a comprehensive insight into the biological state and behavioral mechanisms of cells through the integration of genomics, transcriptomics, epigenomics and proteomics information. Researchers are actively working to further improve the cost-effectiveness, stability and high-throughput capabilities of single-cell multi-omics sequencing technologies and exploring their potential in precision medicine through clinical diagnostics. This review aims to survey the cutting-edge advancements in single-cell multi-omics sequencing, summarizing the representative technologies and their applications in profiling complex diseases, with a particular focus on tumors.

Language models have shown the capacity to learn complex molecular distributions. In the field of molecular generation, they are designed to explore the distribution of molecules, and previous studies have demonstrated their ability to learn molecule sequences. In the early times, recurrent neural networks (RNNs) were widely used for feature extraction from sequence data and have been used for various molecule generation tasks. In recent years, the attention mechanism for sequence data has become popular. It captures the underlying relationships between words and is widely applied to language models. The Transformer-Layer, a model based on a self-attentive mechanism, also shines the same as the RNN-based model. In this research, we investigated the difference between RNNs and the Transformer-Layer to learn a more complex distribution of molecules. For this purpose, we experimented with three different generative tasks: the distributions of molecules with elevated scores of penalized LogP, multimodal distributions of molecules and the largest molecules in PubChem. We evaluated the models on molecular properties, basic metrics, Tanimoto similarity, etc. In addition, we applied two different representations of the molecule, SMILES and SELFIES. The results show that the two language models can learn complex molecular distributions and SMILES-based representation has better performance than SELFIES. The choice between RNNs and the Transformer-Layer needs to be based on the characteristics of dataset. RNNs work better on data focus on local features and decreases with multidistribution data, while the Transformer-Layer is more suitable when meeting molecular with larger weights and focusing on global features.

We identified 11 conserved stretches in over 6.3 million SARS-CoV-2 genomes including all the major variants of concerns. Each conserved stretch is ≥100 nucleotides in length with ≥99.9% conservation at each nucleotide position. Interestingly, six of the eight conserved stretches in ORF1ab overlapped significantly with well-folded experimentally verified RNA secondary structures. Furthermore, two of the conserved stretches were mapped to regions within the S2-subunit that undergo dynamic structural rearrangements during viral fusion. In addition, the conserved stretches were significantly depleted for zinc-finger antiviral protein (ZAP) binding sites, which facilitated the recognition and degradation of viral RNA. These highly conserved stretches in the SARS-CoV-2 genome were poorly conserved at the nucleotide level among closely related β-coronaviruses, thus representing ideal targets for highly specific and discriminatory diagnostic assays. Our findings highlight the role of structural constraints at both RNA and protein levels that contribute to the sequence conservation of specific genomic regions in SARS-CoV-2.

Single-cell clustering is the most significant part of single-cell RNA sequencing (scRNA-seq) data analysis. One main issue facing the scRNA-seq data is noise and sparsity, which poses a great challenge for the advance of high-precision clustering algorithms. This study adopts cellular markers to identify differences between cells, which contributes to feature extraction of single cells. In this work, we propose a high-precision single-cell clustering algorithm-SCMcluster (single-cell cluster using marker genes). This algorithm integrates two cell marker databases(CellMarker database and PanglaoDB database) with scRNA-seq data for feature extraction and constructs an ensemble clustering model based on the consensus matrix. We test the efficiency of this algorithm and compare it with other eight popular clustering algorithms on two scRNA-seq datasets derived from human and mouse tissues, respectively. The experimental results show that SCMcluster outperforms the existing methods in both feature extraction and clustering performance. The source code of SCMcluster is available for free at https://github.com/HaoWuLab-Bioinformatics/SCMcluster.

The expression and activity of transcription factors, which directly mediate gene transcription, are strictly regulated to control numerous normal cellular processes. In cancer, transcription factor activity is often dysregulated, resulting in abnormal expression of genes related to tumorigenesis and development. The carcinogenicity of transcription factors can be reduced through targeted therapy. However, most studies on the pathogenic and drug-resistant mechanisms of ovarian cancer have focused on the expression and signaling pathways of individual transcription factors. To improve the prognosis and treatment of patients with ovarian cancer, multiple transcription factors should be evaluated simultaneously to determine the effects of their protein activity on drug therapies. In this study, the transcription factor activity of ovarian cancer samples was inferred from virtual inference of protein activity by enriched regulon algorithm using mRNA expression data. Patients were clustered according to their transcription factor protein activities to investigate the association of transcription factor activities of different subtypes with prognosis and drug sensitivity for filtering subtype-specific drugs. Meanwhile, master regulator analysis was utilized to identify master regulators of differential protein activity between clustering subtypes, thereby identifying transcription factors associated with prognosis and assessing their potential as therapeutic targets. Master regulator risk scores were then constructed for guiding patients' clinical treatment, providing new insights into the treatment of ovarian cancer at the level of transcriptional regulation.

Moraxella catarrhalis is a symbiotic as well as mucosal infection-causing bacterium unique to humans. Currently, it is considered as one of the leading factors of acute middle ear infection in children. As M. catarrhalis is resistant to multiple drugs, the treatment is unsuccessful; therefore, innovative and forward-thinking approaches are required to combat the problem of antimicrobial resistance (AMR). To better comprehend the numerous processes that lead to antibiotic resistance in M. catarrhalis, we have adopted a computational method in this study. From the NCBI-Genome database, we investigated 12 strains of M. catarrhalis. We explored the interaction network comprising 74 antimicrobial-resistant genes found by analyzing M. catarrhalis bacterial strains. Moreover, to elucidate the molecular mechanism of the AMR system, clustering and the functional enrichment analysis were assessed employing AMR gene interactions networks. According to the findings of our assessment, the majority of the genes in the network were involved in antibiotic inactivation; antibiotic target replacement, alteration and antibiotic efflux pump processes. They exhibit resistance to several antibiotics, such as isoniazid, ethionamide, cycloserine, fosfomycin, triclosan, etc. Additionally, rpoB, atpA, fusA, groEL and rpoL have the highest frequency of relevant interactors in the interaction network and are therefore regarded as the hub nodes. These genes can be exploited to create novel medications by serving as possible therapeutic targets. Finally, we believe that our findings could be useful to advance knowledge of the AMR system present in M. catarrhalis.