P07-29
Automated Hit-to-Lead Optimization Using the SINCHO Protocol and ChemTS
Genki KUDO *1, Shota NAKAJIMA2, Yudai ICHIKAWA3, Takumi HIRAO4, Ryunosuke YOSHINO4, 5, Hitoshi KAMIJIMA2, Takatsugu HIROKAWA4, 5
1Pure and Appl. Sci., Grad. Sci. Tech., University of Tsukuba
2Research Institute of Systems Planning, Inc.
3Med., Med. and Health Sci., University of Tsukuba
4Faculty Med., University of Tsukuba
5TMRC., University of Tsukuba
( * E-mail: s2330052@u.tsukuba.ac.jp)
The Hit-to-lead process in drug discovery and development involves optimizing a hit compound, which initially has low affinity and selectivity, into a lead compound with high affinity and selectivity. For the rational lead compound design, the 3D structural information of the target protein is crucial. This information has become more accessible with the advent of AlphaFold technology and advances in crystal structure analysis. Nevertheless, in the structure-based drug design, the utilization of this valuable information is limited because the hit-to-lead process still heavily relies on the trial-and-error approach of medicinal chemists. Therefore, a computational method to support and replace this manual process is needed.
In this study, we introduce an automatic hit-to-lead system using SINCHO protocol [1,2], which we developed, and ChemTS [3], a de novo molecular generator. This system begins with the 3D structure of a protein–hit compound complex. Based on the structure, the SINCHO protocol identifies the protein pocket and R-point pair that have the potential to improve affinity through substructure modification. The appropriate molecular weight and logP for the modified substructure are also predicted from the SINCHO results. Then, the lead compound candidates are designed using ChemTS, aligning with these predicted properties.
We will provide details of this system and present the case study.
Reference
[1] Kudo G, et al. J. Chem. Inf. Model. 2024;64(11):4475-4484.
[2] Kudo G, et al. Bioinformatics. 2023;39(4):btad212.
[3] Yang X, et al. Sci. Technol. Adv. Mater. 2017;18(1):972-976.