A 3D Structural Interactome to Explore the Impact of Evolutionary Divergence, Population Variation and Small-molecule Drugs on Protein-Protein Interactions Between SARS-CoV-2 and Humans. SD Wierbowski, S Liang, Y Chen, NM Andre, SM Lipkin, GR Whittaker, H Yusubmitted for peer review. 2020. Preprint available on bioRxiv.


3D-SARS2 is a comprehensive data resource to explore protein-protein interactions between SARS-CoV-2 and humans at a structural level. Leveraging state-of-the-art interface prediction (Meyer et al 2018) structural modelling, and docking protocols we provide interface annotations for all published SARS-CoV-2-Human interactions at up to atomic resolution. Our initial release includes 332 interactions reported by Gordon et al 2020 and will constantly be updated as additional interactions are reported.

To facilitate further study of the impact and importance of these interactions, SARS-CoV-2-Human Interactome Browser provides four main analyses. First annotations of the interface residues for each interaction are predicted using our previously published ECLAIR (Ensemble Classifier Learning Algorithm to predict Interface Residues) framework and where possible further refined into atomic resolution structures using a guided docking protocol. These structures and interface annotations are available to view interactively. Second to explore signals of evolution on these interactions we curate a list of all mutations (either viral variations between SARS-CoV and SARS-CoV-2 or human population variants reported by gnomAD) on each interacting protein and report the enrichment for mutation along the interaction interface. Third to explore the impact of these mutations on interaction stability and identify residues most important to the interaction, we perform in-silico scanning mutagenesis and report the predicted change in binding affinity (ΔΔG) for all possible mutations along the interface. For the viral protein we additionally report the overall predicted change in binding affinity for the cumulative set of mutations from SARS-CoV to SARS-CoV-2. Finally, to explore potential drug repurposing to target SARS-CoV-2-Human interactions we provide docked structures for any drugs known to target human interactors of SARS-CoV-2 to compare the drug and protein binding sites.


Version 1,0 July 2020
Initial release including 332 SARS-CoV-2-Human protein-protein interactions reported by Gordon et al 2020.


SARS-CoV-2-Human Interactome Browser is provided by the Yu lab at Cornell University. Please contact Shayne Wierbowski at for any issues or additional data requests. For issues with the website, please contact Charles Liang at Please also feel free to suggest any new publications for which interactions should be added in future updates.