Integrated In silico small molecule drug repurposing approach targeting KRAS G12D therapies for pancreatic ductal adenocarcinoma

Background Pancreatic ductal adenocarcinoma (PDAC) is among the most lethal cancers and is primarily driven by KRAS point mutations, particularly KRAS G12D. MRTX1133, a selective noncovalent inhibitor of KRAS G12D, was discontinued in its phase II clinical development due to its suboptimal pharmacokinetic properties. In this study, we employed an integrated in silico repurposing and screening pipeline that incorporates molecular docking and molecular dynamics simulations to identify promising alternative inhibitors of PDAC. Methodology The study examined 17,430 compounds from DrugBank, and 12,683 unique molecules were filtered based on Lipinski’s rules and Veber’s law. LIGHTHOUSE, an AI-based drug discovery platform, was utilised to predict compound-KRAS G12D interaction scores and IC 50 values for high-throughput screening. 352 high-confidence candidates were selected based on their interaction scores ( ≥ 6.2) and IC 50 ( < 1 µM). The switch II pocket of KRAS G12D was mainly focused on the Glide module in molecular docking as a prominent druggable site. Six (Drugs A-F; five investigational drugs (highest docking score) and MRTX1133) were shortlisted for molecular dynamics refinement. Results Among five experimental drugs, Drug E (DB07793) was shown to possess the best drug-likeness, the most stable interaction profile, and ADMET profiling. The combined computational pipeline highlights that Drug E surmounts MRTX1133 as a repurposed KRAS G12D inhibitor, proposed for PDAC. Significance Statement We successfully screened Drug E (DB07793) as a pharmacologically promising candidate for PDAC. Further validation via in vitro and in vivo studies could establish Drug E as a potential inhibitor of KRAS G12D-driven pancreatic cancer.