Knowledge Discovery and Drug-Repurposing Framework for Pancreatic Ductal Adenocarcinoma: Molecular Networking and Computational Docking

Pancreatic ductal adenocarcinoma (PDAC) remains one of the most lethal malignancies, driven by profound molecular heterogeneity and resistance to current therapy. To support systematic target identification, we established a proteomics-anchored knowledge discovery framework integrating cross-model proteomics harmonization, network topology, high-confidence structural modeling, and large-scale in silico docking. From 1,975 proteins consistently detected across murine and human PDAC models, 32 immunohistochemically confirmed candidates were prioritized for structure-based screening against 7,509 clinically characterized compounds. Blind docking, refined pose sampling, ligand-efficiency scoring, and ADME filtering identified EIF2A, STAM, ANXA2, and AHNAK2 as robustly druggable targets. These proteins exhibited high-affinity interactions with zavegepant (a clinically approved CGRP receptor antagonist), omilancor, bemcentinib, conivaptan, and APTO-253. Docking validation (RMSD 1.98 to 2.56 Å) confirmed methodological reliability, and network analyses placed the 4 proteins within modules linked to endosomal/membrane trafficking and invasive phenotypes. Survival analyses in 176 PDAC patients further supported their clinical relevance. Thus, we suggest a systems-level platform for nominating ligandable PDAC targets and clinically actionable compounds. The framework highlights opportunities for rational drug repurposing and motivates future mechanistic studies at the intersection of proteomics and structure-based screening for targets to PDAC.