SIMPLAIX Project 8: Machine-Learning Aided Multiscale Modeling of Interprotein Electron Transfer and Enzyme Mechanism

2.06.2026

Cytochrome P450 enzymes (CYPs) are membrane-bound monooxygenases that play central roles in drug metabolism and steroid biosynthesis. They are able to process a wide range of chemically diverse ligands. Catalysis requires the transfer of electrons from partner redox proteins and is modulated by homo- and hetero-oligomerization in the membrane. However, the molecular and structural determinants of electron transfer, ligand binding and protein–protein complex formation remain incompletely understood.

This project investigates the mechanisms influencing electron transfer and catalytic activity in CYP systems using a combined approach of multiscale molecular modeling and machine learning. Here, machine learning is applied in two complementary ways. First, machine-learned force fields trained on quantum-mechanical reference data will enable efficient molecular simulations. These simulations will facilitate the study of conformational dynamics, ligand binding and protein complexation of CYP systems within membrane environments. Second, machine learning models will be developed to predict electron transfer rates, which are often the rate-limiting steps of CYP catalysis, directly from protein sequence and available experimental kinetic data.