克拉斯
生物信息学
计算生物学
化学
共价键
小分子
脚手架
生物化学
共价结合
癌症研究
威罗菲尼
组合化学
突变
标杆管理
药物发现
对接(动物)
分区(防火)
计算机科学
药理学
生物信息学
连接器
生物
分子动力学
分子模型
数量结构-活动关系
生物物理学
作者
Glen J. Weiss,Joseph C. Loftus,David Mallery,Nhan L. Tran
出处
期刊:Cancers
[Multidisciplinary Digital Publishing Institute]
日期:2026-04-25
卷期号:18 (9): 1367-1367
标识
DOI:10.3390/cancers18091367
摘要
Background/Objectives: KRAS G12C mutations define a clinically actionable subset of solid tumors, particularly non–small cell lung cancer. Clinical responses to approved covalent inhibitors remain limited by intrinsic and acquired resistance, highlighting the need for structurally distinct inhibitor scaffolds to expand therapeutic options. The objective of this study was to identify novel covalent binders targeting the KRAS G12C switch-II pocket through large-scale in silico screening and experimental validation. Methods: More than 1.9 million small molecules from diverse commercial libraries were screened using covalent docking, followed by multi-stage refinement incorporating molecular dynamics simulations, MM/GBSA free-energy estimation, and cancer-focused QSAR modeling. Results: This integrated workflow yielded 50 prioritized compounds spanning several chemically distinct scaffold classes. These candidates displayed favorable predicted binding energetics, stable ligand-protein interactions over extended simulation timescales, and low structural similarity to clinically approved KRAS G12C inhibitors sotorasib and adagrasib. Benchmarking against these clinical agents, using identical computational parameters, yielded comparable predicted binding energies for several candidate molecules. In cellular NanoBRET target-engagement assays, selected scaffolds, including K788-7251 and AN-989/14669131, exhibited sub-micromolar engagement of KRAS G12C with minimal endothelial cytotoxicity. Conclusions: Collectively, these findings identify structurally distinct, KRAS G12C inhibitor chemotypes and provide tractable starting points for the development of next-generation targeted therapies.
科研通智能强力驱动
Strongly Powered by AbleSci AI