Fig. 2: Top docking poses of dipyridamole and its QMO-optimized variant with SARS-CoV-2 Mpro, obtained using AutoDock Vina.
From: Optimizing molecules using efficient queries from property evaluations
a–d, Top docking poses of dipyridamole (a) and its QMO-optimized variant (b) and their 2D structures (c,d). QMO optimizes the predicted affinity for the dipyridamole variant from 3.94 to 7.59, while maintaining a Tanimoto similarity score of 0.58 and without changing the binding pocket substantially. MM/PBSA calculations for these poses show an improvement in binding free energy from −11.49 kcal mol−1 to −25.65 kcal mol−1. Important residues from the Mpro substrate-binding pocket are also shown. Details are provided in Supplementary Table 2.
