MONSTROUS: a web-based chemical-transporter interaction profiler

• DOI: 10.3389/fphar.2025.1498945
• Publication Date: 2025-02-26T05:13:12Z
• Authors: Mohamed Diwan M. ...; Souvik Dey; Zhen Xu; Ben Clancy; Valmik Desai; Anders Wallqvist
• Keywords: ABC transporters; transporter profiler; SLC transporters; graph convolutional neural network; transporter screening; Therapeutics. Pharmacology; RM1-950; chemical transporter interactions

Transporters are membrane proteins that critical for normal cellular function and mediate the transport of endogenous exogenous chemicals. Chemical interactions with these transporters have potential to affect pharmacokinetic properties drugs. Inhibition can cause adverse drug-drug toxicity, whereas if a drug is substrate transporter, it could lead reduced therapeutic effects. The importance in efficacy toxicity has led regulatory agencies, such as U.S. Food Drug Administration European Medicines Agency, recommend screening new molecular entities transporter interactions. To aid rapid prioritization candidates without liability, we developed publicly available, web-based profiler, MOlecular traNSporT inhibitoR predictOr Utility Server (MONSTROUS), predicts chemical interact recommended testing by agencies. We utilized available data machine learning or similarity-based classification models predict inhibitors substrates 12 transporters. used graph convolutional neural networks (GCNNs) develop predictive sufficient bioactivity data, implemented two-dimensional approach those data. GCNN inhibitor an average five-fold cross-validated receiver operating characteristic area under curve (ROC-AUC) 0.85 ± 0.07, ROC-AUC 0.79 0.08. along applicability domain calculations easy-to-use web interface made at https://monstrous.bhsai.org/.