Swapnil Chavan obtained his Master’s degree in Pharmaco‑informatics from NIPER, India, and earned his PhD in Computational Chemistry from Linnaeus University, Sweden. He subsequently conducted postdoctoral research at the National Center for Computational Toxicology (NCCT) within the U.S. Environmental Protection Agency. He is currently working as a Computational Toxicologist at RISE AB, Sweden.
Research Interests
Swapnil Chavan conducts research at the intersection of computational toxicology, artificial intelligence, and cheminformatics, with a focus on developing next-generation approaches for chemical hazard and risk assessment. His work spans predictive metabolism, hazard prediction, explainable AI, neurosymbolic AI, and agentic AI, with the goal of creating transparent, reliable, and scientifically grounded decision-support systems for chemical safety evaluation.
OpenTox Summer School 2026
Why Is Predicting Drug Metabolism Still So Hard? Challenges and AI-Based Solutions for Metabolite Prediction
Predicting drug metabolism remains one of the most important problems in computational toxicology and drug development. Although advances in artificial intelligence have significantly improved predictive modeling, accurate prediction of metabolic transformations remains challenging. In this talk, I will present METABOLITE-GEN, a graph-based deep learning framework that integrates site-of-metabolism prediction, reaction class prediction, and metabolite generation within a unified workflow. The talk will discuss the development and evaluation of the approach, highlighting the opportunities and limitations of modern AI methods and underscoring the importance of high-quality data and robust model development for advancing predictive metabolism and computational toxicology.