End-Point Affinity Estimation of Galectin Ligands by Classical and Semiempirical Quantum Mechanical Potentials

• DOI: 10.1021/acs.jcim.4c01659
• Publication Date: 2025-01-04T14:28:16Z
• Authors: Jan Choutka; Jakub Kaminský; Ercheng Wang; Kamil Parkan; Radek Pohl
• Keywords: Affinities; Implicit solvation

The use of quantum mechanical potentials in protein–ligand affinity prediction is becoming increasingly feasible with growing computational power. To move forward, validation such on real-world challenges necessary. this end, we have collated an extensive set over a thousand galectin inhibitors known affinities and docked them into galectin-3. poses were then used to systematically evaluate several modern force fields semiempirical (SQM) methods up the tight-binding level under consistent workflow. Implicit solvation models available tested simulate effects. Overall, best study achieved Pearson correlation 0.7–0.8 between computed experimental affinities. There differences their ability rank ligands across entire ligand as well within subsets structurally similar ligands. A major discrepancy was observed for subset that bind protein via halogen bond, which clearly challenging all methods. inclusion entropic term calculated by rigid-rotor-harmonic-oscillator approximation at SQM slightly worsened experiment but brought closer values. We also found success strongly depended model. Furthermore, provide in-depth analysis individual energy terms effect overall accuracy.