Clustering Protein Binding Pockets and Identifying Potential Drug Interactions: A Novel Ligand-Based Featurization Method

• DOI: 10.1021/acs.jcim.3c00722
• Publication Date: 2023-10-17T15:57:41Z
• Authors: Garrett A. Stevenson; Dan Kirshner; Brian J. Bennion; Yue Yang; Xiaohua Zhang; Adam Zemla; Marisa W. Torres; Aidan Epstein; Derek Jones; Hyojin Kim; W. F. Drew Bennett; Sergio E. Wong; Jonathan E. Allen; Felice C. Lightstone
• Keywords: Interpretability; Proteome; Human proteome project; Protein ligand; Drug target

Protein–ligand interactions are essential to drug discovery and development efforts. Desirable on-target or multitarget the first step in finding an effective therapeutic, while undesirable off-target assessing safety. In this work, we introduce a novel ligand-based featurization mapping of human protein pockets identify closely related targets project drugs into hybrid protein–ligand feature space their likely interactions. Using structure-based template matches from PDB, featured by ligands that bind best co-complex matches. The simplicity interpretability approach provide granular characterization proteome at protein-pocket level instead traditional protein-level family, function, pathway. We demonstrate power method clustering subset evaluating predicted cluster associations over 7000 compounds.