Substrate-specificity of pharmacogenomic variability - clinical relevance and emerging predictive strategies

Genetic variation in drug-metabolizing enzymes and transporters contributes to interindividual differences in drug efficacy and toxicity. While current pharmacogenomic practice typically assigns fixed functional categories to variant alleles, accumulating evidence shows that many variant alleles exert substrate-specific effects.

We here summarize the current evidence for substrate-dependent functional variability across key pharmacogenes, including CYP2D6, CYP2C8, CYP2C9 and multiple uptake and efflux transporters, based on both experimental in vitro approaches and clinical pharmacokinetic analyses. We further evaluate the performance of existing computational variant-effect predictors and outline emerging structural, sequence-based, and representation-learning strategies for the prediction of substrate-specificity. Particular focus is placed on methodological gaps that currently prevent the prediction of substrate-specific effects and on the evolving technological landscape that may enable such capabilities.

Substrate specificity of pharmacogenetic variations is clinically relevant and insufficiently captured by current functional annotations and prediction tools. We believe that new strategies based on integrative multimodal frameworks that combine structural descriptors, protein language-model embeddings, ligand features and high-quality experimental data have the potential to capture substrate-specific variant effects in the near future. These advances will provide exciting new opportunities to systematically evaluate variant function, which will benefit both drug development and precision medicine.