Berlin 2024 – scientific programme

Parts | Days | Selection | Search | Updates | Downloads | Help

BP: Fachverband Biologische Physik

BP 10: Computational Biophysics II

BP 10.9: Talk

Tuesday, March 19, 2024, 12:00–12:15, H 0112

Enhancing protein-ligand binding affinity via optimal selection of water molecules — •Miljan Dašić, Jindřich Fanfrlík, and Jan Řezáč — Institute of Organic Chemistry and Biochemistry of the Czech Academy of Sciences, Flemingovo náměstí 2, 166 10, Prague, Czech Republic

Accurate and fast determination of protein-ligand (P-L) binding affinity represents a foundational problem of computational biophysics. A promising solution is an universal physics-based scoring function SQM2.20 based on semi-empirical quantum-mechanical computational methods. Its performance has been rigorously verified over a benchmark dataset PL-REX consisting of high-resolution crystal structures and trustworthy experimentally determined P-L affinities (10 diverse protein targets; 164 QM-optimized P-L complexes). Presence of water molecules has a significant impact on P-L binding affinity, via formation of hydrogen bond bridges. We have developed a computational tool which optimally selects waters enhancing the P-L binding affinity. Each of the ten protein targets comprises different crystals. Waters present in all of them represent the input for selection procedure. Such procedure includes clustering and comparison of waters contained in clusters with waters present in one selected reference crystal. Presence of optimally selected waters improves the correlation with experimental data. We investigated the sensitivity of scoring on the geometry of crystals. For each protein target, we determined the best reference crystal which maximizes the scoring results.

Keywords: protein-ligand binding affinity; semiempirical quantum-mechanical methods; physics-based scoring function; PL-REX benchmark dataset; impact of water molecules