The discovery and repurposing of drugs require a deep understanding of the mechanism of drug action (MODA). Existing computational methods mainly model MODA with the protein-protein interaction (PPI) network. However, the molecular interactions of drugs in the human body are far beyond PPIs. Additionally, the lack of interpretability of these models hinders their practicability. We propose an interpretable deep learning-based path-reasoning framework (iDPath) for drug discovery and repurposing by capturing MODA on by far the most comprehensive multilayer biological network consisting of the complex high-dimensional molecular interactions between genes, proteins and chemicals. Experiments show that iDPath outperforms state-of-the-art machine learning methods on a general drug repurposing task. Further investigations demonstrate that iDPath can identify explicit critical paths that are consistent with clinical evidence. To demonstrate the practical value of iDPath, we apply it to the identification of potential drugs for treating prostate cancer and hypertension. Results show that iDPath can discover new FDA-approved drugs. This research provides a novel interpretable artificial intelligence perspective on drug discovery.

Coronavirus disease 2019 (COVID-19) has spread rapidly worldwide, causing significant mortality. There is a mechanistic relationship between intracellular coronavirus replication and deregulated autophagosome-lysosome system. We performed transcriptome analysis of peripheral blood mononuclear cells (PBMCs) from COVID-19 patients and identified the aberrant upregulation of genes in the lysosome pathway. We further determined the capability of two circulating markers, namely microtubule-associated proteins 1A/1B light chain 3B (LC3B) and (p62/SQSTM1) p62, both of which depend on lysosome for degradation, in predicting the emergence of moderate-to-severe disease in COVID-19 patients requiring hospitalization for supplemental oxygen therapy. Logistic regression analyses showed that LC3B was associated with moderate-to-severe COVID-19, independent of age, sex and clinical risk score. A decrease in LC3B concentration <5.5 ng/ml increased the risk of oxygen and ventilatory requirement (adjusted odds ratio: 4.6; 95% CI: 1.1-22.0; P = 0.04). Serum concentrations of p62 in the moderate-to-severe group were significantly lower in patients aged 50 or below. In conclusion, lysosome function is deregulated in PBMCs isolated from COVID-19 patients, and the related biomarker LC3B may serve as a novel tool for stratifying patients with moderate-to-severe COVID-19 from those with asymptomatic or mild disease. COVID-19 patients with a decrease in LC3B concentration <5.5 ng/ml will require early hospital admission for supplemental oxygen therapy and other respiratory support.

Tumor suppressor genes (TSGs) exhibit distinct evolutionary features. We speculated that TSG promoters could have evolved specific features that facilitate their tumor-suppressing functions. We found that the promoter CpG dinucleotide frequencies of TSGs are significantly higher than that of non-cancer genes across vertebrate genomes, and positively correlated with gene expression across tissue types. The promoter CpG dinucleotide frequencies of all genes gradually increase with gene age, for which young TSGs have been subject to a stronger evolutionary pressure. Transcription-related features, namely chromatin accessibility, methylation and ZNF263-, SP1-, E2F4- and SP2-binding elements, are associated with gene expression. Moreover, higher promoter CpG dinucleotide frequencies and chromatin accessibility are positively associated with the ability of TSGs to resist downregulation during tumorigenesis. These results were successfully validated with independent datasets. In conclusion, TSGs evolved specific promoter features that optimized cancer resistance through achieving high expression in normal tissues and resistance to downregulation during tumorigenesis.

Patients with spinal muscular atrophy (SMA) are susceptible to the respiratory infections and might be at a heightened risk of poor clinical outcomes upon contracting coronavirus disease 2019 (COVID-19). In the face of the COVID-19 pandemic, the potential associations of SMA with the susceptibility to and prognostication of COVID-19 need to be clarified. We documented an SMA case who contracted COVID-19 but only developed mild-to-moderate clinical and radiological manifestations of pneumonia, which were relieved by a combined antiviral and supportive treatment. We then reviewed a cohort of patients with SMA who had been living in the Hubei province since November 2019, among which the only 1 out of 56 was diagnosed with COVID-19 (1.79%, 1/56). Bioinformatic analysis was carried out to delineate the potential genetic crosstalk between SMN1 (mutation of which leads to SMA) and COVID-19/lung injury-associated pathways. Protein-protein interaction analysis by STRING suggested that loss-of-function of SMN1 might modulate COVID-19 pathogenesis through CFTR, CXCL8, TNF and ACE. Expression quantitative trait loci analysis also revealed a link between SMN1 and ACE2, despite low-confidence protein-protein interactions as suggested by STRING. This bioinformatic analysis could give hint on why SMA might not necessarily lead to poor outcomes in patients with COVID-19.

OBJECTIVE: To develop an end-to-end deep learning framework based on a protein-protein interaction (PPI) network to make synergistic anticancer drug combination predictions. MATERIALS AND METHODS: We propose a deep learning framework named Graph Convolutional Network for Drug Synergy (GraphSynergy). GraphSynergy adapts a spatial-based Graph Convolutional Network component to encode the high-order topological relationships in the PPI network of protein modules targeted by a pair of drugs, as well as the protein modules associated with a specific cancer cell line. The pharmacological effects of drug combinations are explicitly evaluated by their therapy and toxicity scores. An attention component is also introduced in GraphSynergy, which aims to capture the pivotal proteins that play a part in both PPI network and biomolecular interactions between drug combinations and cancer cell lines. RESULTS: GraphSynergy outperforms the classic and state-of-the-art models in predicting synergistic drug combinations on the 2 latest drug combination datasets. Specifically, GraphSynergy achieves accuracy values of 0.7553 (11.94% improvement compared to DeepSynergy, the latest published drug combination prediction algorithm) and 0.7557 (10.95% improvement compared to DeepSynergy) on DrugCombDB and Oncology-Screen datasets, respectively. Furthermore, the proteins allocated with high contribution weights during the training of GraphSynergy are proved to play a role in view of molecular functions and biological processes, such as transcription and transcription regulation. CONCLUSION: The introduction of topological relations between drug combination and cell line within the PPI network can significantly improve the capability of synergistic drug combination identification.

Journal of the American Medical Informatics Association, ocab162, https://doi.org/10.1093/jamia/ocab162 When this paper was first published, there were a number of formatting problems in the pdf version of the paper, especially affecting the equations. The html version of the paper was correct. These errors were not due to author error, and the publisher apologizes for these errors. These errors have now been corrected online.

MOTIVATION: Increasing amounts of whole exome or genome sequencing data present the challenge of analysing rare variants with extremely small minor allele frequencies. Various statistical tests have been proposed, which are specifically configured to increase power for rare variants by conducting the test within a certain bin, such as a gene or a pathway. However, a gene may contain from several to thousands of markers, and not all of them are related to the phenotype. Combining functional and non-functional variants in an arbitrary genomic region could impair the testing power. RESULTS: We propose a Zoom-Focus algorithm (ZFA) to locate the optimal testing region within a given genomic region. It can be applied as a wrapper function in existing rare variant association tests to increase testing power. The algorithm consists of two steps. In the first step, Zooming, a given genomic region is partitioned by an order of two, and the best partition is located. In the second step, Focusing, the boundaries of the zoomed region are refined. Simulation studies showed that ZFA substantially increased the statistical power of rare variants' tests, including the SKAT, SKAT-O, burden test and the W-test. The algorithm was applied on real exome sequencing data of hypertensive disorder, and identified biologically relevant genetic markers to metabolic disorders that were undetectable by a gene-based method. The proposed algorithm is an efficient and powerful tool to enhance the power of association study for whole exome or genome sequencing data. AVAILABILITY AND IMPLEMENTATION: The ZFA software is available at: http://www2.ccrb.cuhk.edu.hk/statgene/software.html. CONTACT: [email protected] or [email protected]. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.