The Second CACHE Challenge - Targeting the RNA-Binding Pocket of the SARS-CoV2 Nonstructural Protein 13 via a consensus-scoring method and FITTED templated docking.

Disrupting the Nonstructural Protein 13 (NSP13) in SARS-CoV2 could provide a great avenue for the treatment of COVID-19 and help reduce its enormous health burden. As part of the second CACHE challenge, we targeted each of two sub-pockets of the NSP13 RNA-binding site via a multi-pronged virtual screening (VS) campaign, using the latest functionality in FITTED, our docking program, part of the FORECASTER drug discovery suite. After extensive structure preparation and docking (rigid, flexible), we evaluated predicted poses from the VS using four approaches: docking score, machine learning (graph neural network), quantum-mechanics, and visualization, with the final selection being based on the consensus of all four approaches. Additionally, we implemented templated docking within FITTED to take advantage of fragments co-crystallized with NSP13, which supplemented our consensus selection. We now await the experimental testing of our predictions by the Structural Genomics Consortium, and once available, we will update this manuscript accordingly. In sharing our approach and findings, we hope to continue contributing to open science, and engaging in the ongoing effort of the scientific community towards ending COVID-19.