Quantum Mechanics-Based Ranking of Predicted Proteolysis Targeting Chimeras-Mediated Ternary Complexes

Abstract

Targeted protein degradation has become the most pursued alternative modality to small-molecule inhibition over the past decade. The traditional strategy of blocking protein activity by tightly binding to a functional substrate pocket has progressed toward proteolysis-targeting chimeras (PROTACs), bivalent molecules that induce the knockdown of targeted proteins. Herein, a combined protocol is described for modeling ternary complexes via well-established approaches. We performed local protein–protein docking using Rosetta protocol and sampled the conformational landscape of a specific PROTAC molecule that was compatible with the generated protein–protein docking poses, followed by double and independent single-linkage/nearest-neighbor clustering for representative selection. Subsequently, we combined the fragment molecular orbital and density functional tight-binding methods to facilitate fast quantum mechanics-based energy calculations of the clustered ternary complexes. Finally, the computed energy values were utilized to score and select the best ternary poses, achieving good agreement with available crystallographic data.