In Silico Drug Repurposing: Integrating Network Pharmacology and Structure-Based Virtual Screening to Accelerate Preclinical Drug Development

De novo drug discovery requires over a decade of research and $1.5 billion to $2.5 billion in investment per approved molecule, yet more than 90% of candidates fail during clinical trials due to unforeseen toxicity or poor efficacy. In silico drug repurposing offers a faster, cost-effective alternative by finding new therapeutic uses for regulatory-approved or clinical-stage molecules. However, individual computational methods have clear blind spots. Network biology identifies disease-related genes but lacks atomic-level detail on how a drug physically binds to a receptor. Conversely, standalone molecular docking often produces false-positive hits and relies on an outdated single-target mindset. This review provides a practical blueprint for combining network pharmacology with structure-based virtual screening (SBVS) in preclinical drug repurposing. We first outline how disease genomics databases (DisGeNET, GeneCards) and protein–protein interaction (PPI) networks (STRING) pair with graph-theory tools in Cytoscape (cytoHubba, MCODE) to pinpoint central hub proteins. Next, we detail how to screen curated chemical libraries (DrugBank, ChEMBL) against these prioritized targets using AutoDock Vina and AlphaFold structures. We also explain why approved drugs still require indication-specific pharmacokinetic checks in SwissADME and ADMETlab 3.0, followed by 100–500 ns explicit-solvent molecular dynamics (MD) simulations in GROMACS and MM-PBSA/GBSA free-energy rescoring. Finally, we review recent preclinical applications across infectious, neurological, and oncological diseases and discuss how deep-learning tools like DiffDock and AlphaFold 3 are reshaping virtual screening.