DUAL-MODALITY IN SILICO DESIGN OF DIVARASIB- AND GARSORASIB-DERIVED INHIBITORS AND VHL-BASED PROTAC DEGRADERS AGAINST KRAS G12C IN NON-SMALL CELL LUNG CANCER

KRAS G12C is an established oncogenic driver in non-small cell lung cancer (NSCLC), but the covalent inhibitors approved against it often produce only partial and short-lived responses. Using the clinical inhibitors Divarasib and Garsorasib as starting points, we investigated two complementary strategies against this target: direct small-molecule inhibition and PROTAC-mediated degradation. Network pharmacology was first applied to define the signalling context of KRAS G12C. SwissTargetPrediction identified 99 unique targets for the two parent drugs, and overlap with 260 NSCLC-associated genes from Open Targets yielded 39 shared targets. In the resulting STRING network (39 nodes, 480 edges), KRAS ranked first in all six cytoHubba algorithms, followed by the hub genes AKT1, PIK3CA, MTOR, EGFR, BCL2 and HSP90AA1. KEGG enrichment highlighted PI3K-Akt signalling and EGFR tyrosine kinase inhibitor resistance. Two derivative series (D1-D8 from Divarasib and G1-G8 from Garsorasib) and four VHL-recruiting PROTACs were then designed and docked into the switch-II pocket of KRAS G12C (PDB 6OIM). Derivatives D3 and G8 scored approximately 1.0 and 1.2 kcal/mol better than their respective parent drugs, and PROTAC G2, which combines a Garsorasib warhead with a rigid alkyne-containing linker, was the best-performing degrader. SwissADME, pkCSM and ProTox predicted properties beyond the Rule of Five, ProTox toxicity class 5 for the parents and derivatives, and potential hERG II and hepatotoxicity risks. These findings are computational predictions only. Subsequent work should include molecular dynamics simulations, synthesis, and biochemical and cellular KRAS G12C degradation assays (DC50, Dmax). It should also include structure-guided refinement of the linker and exit vector to lower molecular weight and hepatic risk, and evaluation in resistance models driven by the EGFR and PI3K-AKT pathways.