Therapeutic candidates for SARS-CoV-2: in silico predictions and selected experimental validation for public small molecule libraries and designed synthetic antibodies

Results of computationally screening public small-molecule libraries for binding to Covid-19 protein binding sites (four sites on two proteins, initially), including both mechanistic (physics-based) calculations and machine learning-based predictions, along with selected experimental validation. In addition, model predictions of safety and pharmacokinetic parameters are provided for each molecule, as well as molecular descriptors that are features used in the models. Finally, selected designed synthetic antibodies and binding predictions are also available.