ABSTRACT The present study describes the identification of novel covalent inhibitors of the KRAS G12C mutant protein, a key oncogenic driver in multiple cancers. Using a structure‐based approach, a dataset of 40 quinazolinone analogs structurally related to the FDA‐approved inhibitor Sotorasib (AMG 510) was investigated. Covalent molecular docking was performed using the CovDock module, with the cysteine 12 residue (Cys12) designated as the covalent attachment site. Compound 11 was identified as a lead candidate, forming multiple interactions, including a covalent bond with Cys12, hydrogen bonds with Glu63, and π–π stacking interactions with Tyr96, demonstrating a greater docking affinity than Sotorasib. A 3D‐QSAR model was developed using PLS regression, showing strong predictive capabilities for the training set ( R 2 = 0.95, R 2 CV = 0.63) and the test set ( Q 2 = 0.90). A structure‐based approach also facilitated the virtual screening of the PubChem database to identify additional novel covalent inhibitors of the KRAS G12C . The molecular dynamics (MD) of two inhibitors, compound C1 and Sotorasib, were compared. C1 exhibited enhanced structural stability, characterized by a low RMSD (∼1.8 Å), reduced loop mobility, and robust hydrogen bonding interactions with Glu63, Lys16, and Tyr96. In contrast, Sotorasib displayed greater flexibility (RMSD ∼2.06 Å) and depended on transient hydrophobic interactions. The findings suggest that C1 has potential as a therapeutic candidate for KRAS‐driven cancers. Additionally, based on earlier studies, four new compounds were identified as covalent inhibitors of KRAS G12C , exhibiting strong binding affinity, favorable activity predictions, and good ADMET properties. In conclusion, the integration of 3D‐QSAR, covalent docking, and MD simulations led to the discovery of potent C1‐C4 leads. These findings provide a reliable computational strategy for the development of next‐generation KRAS G12C inhibitors with optimized binding profiles.
Belafriekh et al. (Thu,) studied this question.