Large-Scale Ligand-Based Virtual Screening for SARS-CoV-2 Inhibitors Using Deep Neural Networks
Markus Hofmarcher, Andreas Mayr, Elisabeth Rumetshofer, Peter Ruch, Philipp Renz, Johannes Schimunek, Philipp Seidl, Andreu Vall, Michael Widrich, Sepp Hochreiter, and Günter Klambauer
Due to the current severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic, there is an urgent need for novel therapies and drugs. We conducted a large-scale virtual screening for small molecules that are potential CoV-2 inhibitors.
We used ChemAI, a deep neural network of SmilesLSTM type able to predict a large number of biological outcomes, including binding to targets, inhibitory, and toxic effects. ChemAI has been trained on more than 220 million data points across 3.6 million molecules from three public drug-discovery databases (ChEMBL, ZINC, and PubChem).
With ChemAI, we screened one billion molecules from the ZINC database. The SARS-CoV-2 virus has two main proteases that are critical for its replication: 3C-like protease and Papain Like protease. We ranked the molecules based on their ability to inhibit both proteases by the consensus score, and evaluated their potential toxic effects.
We thus reduced the dataset to the 30,000 top-ranked compounds, which are readily accessible and purchasable via the ZINC database. We now provide these top-ranked compounds as a library for further screening with bioassays. We also screened the DrugBank database and obtained a ranked list of potential drug candidates to inhibit the virus.
SSRN 3561442, 2020-03-23.