Dissertation
Virtual Safety First: from chemical characteristics to human-relevant hazard identification
Currently chemical safety testing is mainly performed on test animals. This thesis explores the use of computational models for predicting chemical hazards.
- Author
- L. Schoenmaker
- Date
- 25 June 2026
- Links
- Thesis in Leiden Repository
It focuses on data curation, machine learning models and tools for sharing such models. We show how in silico models can be used for hazard identification by employing Parkinson’s disease as a case study. In addition to this special attention is also paid to de novo generation techniques and physics-based affinity prediction methods.