Forschungsgebiet

Machine learning (ML) is increasingly being integrated into molecular modeling, enabling scientists to explore data and make predictions that experimentalists can test. Following the structure-function paradigm, I am interested in developing an ML for conformation generation and molecular similarity prediction. Using reliable quantum-mechanical (QM) data, we will train an ML model that can predict multiple conformations for a given molecule. This approach benefits small-molecule modeling by providing 3D structures for more efficient optimization, and identifying potential active molecules based on structural and chemical similarities.

Methodik

Molecular Dynamics Simulation, Structural Biochemistry, Molecular Conformation, Machine Learning