A paper co-authored by Kaho Akaki (Akiyama Laboratory, Institute of Science Tokyo), Yutaka Akiyama (same laboratory), and our lab’s Yanagisawa has been published in the open-access Journal of Chemical Information and Modeling.
This study aimed to improve the accuracy of protein–ligand docking calculations, a key step in structure-based virtual screening (VS). The authors performed mixed-solvent molecular dynamics (MSMD) simulations using four types of probe molecules and used the resulting spatial distributions of probe atoms to derive grid free energies for five atom classifications (XS_types), proposing a method to correct the docking scoring function with this information. VS experiments on nine target proteins showed improved accuracy, with the average EF1% increasing from 6.65 to 7.36, demonstrating that the method contributes to more accurate virtual screening.
Kaho Akaki, Keisuke Yanagisawa, Yutaka Akiyama, "Probe-Atom Distributions Obtained from Mixed-Solvent Molecular Dynamics Improve the Scoring of Docking Calculations", Journal of Chemical Information and Modeling 66, 10971-10981, 2026/8. DOI: 10.1021/acs.jcim.6c00947
(translated by Claude)