Network-based drug repurposing for novel coronavirus 2019-nCoV/SARS-CoV-2

Analysis

Why This Paper Matters

This paper addresses the urgent need for effective treatments against the novel coronavirus 2019-nCoV/SARS-CoV-2 during the early stages of the COVID-19 pandemic. By leveraging drug repurposing, it offers a faster and more cost-effective alternative to de novo drug discovery. The network medicine approach integrates multiple data types—phylogenetic, interactome, and transcriptomic—to systematically identify candidate drugs and combinations, which is particularly valuable when time is critical.

The study's significance lies in its ability to quantify the interplay between viral-host interactions and drug targets within the human protein-protein interaction network. This systems-level perspective goes beyond traditional target-based screening and captures the complexity of host-pathogen interactions. The identification of 16 potential drugs and three combinations provides a prioritized list for experimental validation, potentially accelerating clinical translation.

Technical Contributions

• Phylogenetic Analysis: Compared 15 HCoV whole genomes, revealing 79.7% nucleotide identity between 2019-nCoV and SARS-CoV, with envelope and nucleocapsid proteins being highly conserved (96% and 89.6% identity).
• Network Proximity Analysis: Quantified the distance between drug targets and HCoV-host proteins in the human interactome to prioritize drugs that target disease modules.
• Complementary Exposure Pattern: Introduced a novel concept for drug combinations where two drugs hit the same viral-host subnetwork but target separate neighborhoods, potentially reducing resistance and enhancing efficacy.
• Validation via Enrichment Analyses: Used drug-gene signatures and HCoV-induced transcriptomics data from human cell lines to confirm the biological relevance of prioritized drugs.

Results

The study identified 16 potential anti-HCoV repurposable drugs, including melatonin, mercaptopurine, and sirolimus. Three drug combinations were captured: sirolimus plus dactinomycin, mercaptopurine plus melatonin, and toremifene plus emodin. The phylogenetic analysis showed that 2019-nCoV shares 79.7% nucleotide identity with SARS-CoV, with envelope and nucleocapsid proteins being 96% and 89.6% identical, respectively. These results were validated by enrichment analyses of drug-gene signatures and transcriptomics data, though no quantitative performance metrics (e.g., AUC, precision) are reported in the abstract.

Significance

This work provides a powerful, generalizable framework for rapid drug repurposing against emerging viral pathogens. By integrating network medicine with systems pharmacology, it offers a blueprint for future pandemic responses. The methodology can be adapted to other diseases where host-pathogen interactions are known, potentially reducing the time and cost of therapeutic development. For the AI community, it demonstrates the value of network-based machine learning approaches in biomedical discovery, particularly when data is sparse and time is critical.