Integrated ligand-based 3D-QSAR, pharmacophore-guided virtual screening, target prediction, DFT, docking, molecular dynamics simulations, and AI-assisted retrosynthetic route prediction for the rational design of pyrazole–benzimidazole hybrids agains

Colorectal cancer remains a cause of cancer-related mortality, highlighting the need for targeted therapeutics. In this study, an integrated CADD approach was employed to identify pyrazole–benzimidazole hybrids as potential EGFR inhibitors for CRC. An atom-based 3D-QSAR model (R2 = 0.9700, Q2 = 0.9560), validated by bootstrap analysis, Y-scrambling, and applicability domain assessment, identified the key structural features governing antiproliferative activity. The top-ranked AHRRR1 pharmacophore model, combined with SAR analysis, R-group enumeration, and ChEMBL-assisted molecular design, generated a virtual library of 22,148 novel chemical entities. Pharmacophore-based virtual screening, ADMET evaluation, toxicity prediction, off-target liability assessment, and density functional theory calculations prioritized five compounds. Computational target prediction identified EGFR as the probable molecular target of the lead compound, C-19. Molecular docking demonstrated favourable binding of C-19 within the EGFR active site (docking score = −8.891 kcal/mol), supported by key interactions with Asp831 and Met769. MM/GBSA analysis predicted a binding free energy (−65.00 kcal/mol) slightly superior to that of the reference inhibitor Pelitinib (−64.61 kcal/mol), while 250 ns molecular dynamics simulations confirmed the stability of the C-19–EGFR complex. Overall, C-19 emerged as a promising EGFR-targeted lead candidate, providing a computational basis for its synthesis and experimental validation for colorectal cancer.