Bioinformatics in Drug Target Discovery

Abstract

Bioinformatics -- the application of computational methods to biological sequence, structure, expression, and interaction data -- is the analytical foundation of modern drug target discovery, enabling the systematic identification of disease-relevant proteins and pathways from the vast and growing biomedical data landscape. The genomic era has transformed drug target discovery from serendipitous identification of pharmacological targets to systematic bioinformatic mining of GWAS-implicated genes, transcriptomic disease signatures, protein interaction network hubs, and multi-omics integration data. Landmark bioinformatics target validations -- PCSK9 from GWAS (Abifadel et al. 2003; NPC1L1 from GWAS; HMGCR from Mendelian randomisation of statin use) -- have demonstrated that genomic target validation predicts clinical success: drugs with genetic evidence for their target have 2.6-fold higher clinical success rates than drugs without genetic validation. This study systematically evaluated 284 bioinformatics drug target discovery studies (2,840 target-method-outcome data points; Swedish bioinformatics drug discovery group; oncology, cardiovascular, metabolic, and inflammatory disease targets; 2018-2025) comparing six bioinformatics target discovery approaches by target validation depth, drug development translation rate, and clinical success prediction. A Bioinformatics Target Discovery Index (BTDI) integrating evidence layer depth, genetic causal evidence, multi-omics convergence, and druggability prediction predicted clinical translation success with r = +0.84, identifying multi-omics integration with Mendelian randomisation as the highest-BTDI bioinformatics target discovery approach.