Our goal is to accelerate studies of virus mechanism and to inform development of vaccines, diagnostics, therapeutics, and antibodies against SARS-CoV-2, the coronavirus causing the COVID 2019 pandemic.
We create quantitative models using virus sequence variation to predict mutations and 3D structure- here's what we can give so far:
in silico deep mutation scans
mutations visualized on 3D structures
all data (download) &
predicted 3D contacts for proteins (& RNA soon)
virus-host interactions (TBD)
protein complexes & active sites (TBD)
Uploaded: Mutation effect predictions for SARS-CoV-2 proteins, alignments to homologs in other viruses, visualizations on 3D structures, evolutionarily coupled residues and structure predictions.
In progress: RNA mutation predictions, RNA structure predictions, residues and genes in virus-host interactions.
For each SARS-CoV-2 protein, we predicted mutation effects for all possible amino acid substitutions, and in some cases 3D structure. These predictions are based on models inferred from multiple sequence alignments of proteins available through UniProt as of March 2020 [1,2], using the EVcouplings software [3,4,5,6].