Multi-target computational drug repurposing identifies FDA-approved candidates with micronutrient-dependent efficacy modulation for pancreatic ductal adenocarcinoma: the ARES v5 pipeline

Background: Pancreatic ductal adenocarcinoma (PDAC) remains the most lethal solid tumor, with a 5-year survival rate of 13% and limited therapeutic options. Current standard-of-care regimens, gemcitabine-based and FOLFIRINOX, provide modest survival benefit, and targeted therapies remain restricted to the approximately 7% of patients harboring BRCA1/2 mutations. Drug repurposing offers a strategy to rapidly identify candidates with established safety profiles for PDAC. Methods: We developed ARES v5, a multi-target computational pipeline that screens 333 compounds (FDA-approved drugs, investigational agents, and structural analogs) against six oncogenic targets critical in PDAC signaling: EGFR, KRAS, STAT3, SOS1, HSP90, and JAK2. The pipeline integrates molecular docking (AutoDock Vina 1.2.7), ADMET filtering based on structural alerts (hERG, hepatotoxicity, Ames, BBB), synergy analysis with established PDAC therapies, consensus scoring, and a novel micronutrient-dependent efficacy correction module that adjusts predicted binding based on cofactor dependencies specific to PDAC patient nutritional status. Results: Internal cross-validation with five known inhibitors (Gefitinib, Rucaparib, Axitinib, Sunitinib, Momelotinib) confirmed pipeline accuracy. Sensitivity analysis across three weight configurations demonstrated that Olaparib (rank 1) and Apixaban (rank 2) maintained top-3 FDA-approved positions in all configurations, establishing conclusion robustness. Among 212 ADMET-clean compounds, Olaparib (score_final = 0.8875) and Apixaban (score_final = 0.8787) emerged as the top FDA-approved candidates, each binding 5 of 6 targets with favorable docking energies (best dG: -9.26 and -8.64 kcal/mol, respectively). Both compounds showed high micronutrient-dependent sensitivity (45% score reduction with cofactor depletion), with Mg2+ identified as the critical limiting cofactor. Conclusions: ARES v5 identifies Olaparib and Apixaban as robust multi-target candidates for PDAC. While Olaparib's efficacy is limited to BRCA-mutated patients (7% of PDAC), Apixaban targets conserved oncogenic pathways active in more than 90% of PDAC cases. The finding that Apixaban, a generic anticoagulant available globally for less than 1 EUR per day, shows strong multi-target binding against PDAC-critical kinases warrants urgent experimental validation, particularly given the convergence with Edoxaban's experimentally demonstrated STAT3 suppression via the Factor Xa-PAR2 axis.