In silico transporter modeling and its role in computational toxicology

Presented by a partner of the recently established EU-ToxRisk Commercial Platform for New Approach Methods (NAMs) in Safety Assessment.

Transmembrane transporters not only play a major role in ADME but they are also increasingly linked to toxicity. In this Webinar, Gerhard Ecker will outline computational approaches to predict the transporter interaction profile of compounds in order to minimize the risk of failures in drug development. Methods presented comprise classical machine learning models as well as deep learning approaches. The latter was also used to overcome insufficient size and imbalance of toxicity datasets. The combined use of structure-based methods for the prediction of molecular initiating events and machine learning led to a model for mitochondrial toxicity. Leveraging complex data analysis pursued with KNIME workflows allows the creation of compound-pathway interaction fingerprints linked to hepatotoxicity and cardiotoxicity. Finally, ToxPHACTS, a data-driven tool for toxicological read-across will be presented.