Thesis Talk: Floris Löffler

Thesis Talk
Automated In Silico Prediction of Protein Binding Affinities for Metal Complexes
Floris Löffler
Date
Friday 05 Sep 2025
Time
15:15 - 15:45
Location
BW006
Supervisor
Sylvestre Bonnet
2nd reviewer
Francesco Buda
Jury
Anthe Janssen

Virtual screening (VS) is a computational technique in which a large score of compounds is tested on their ability to inhibit potential drug targets. The use of VS requires computational tools capable of obtaining binding affinities for compounds, such as metal complexes, to proteins. In this work, efforts towards the development and testing of a computational tool adapted from MetalFlow, capable of obtaining such binding affinities for metal complexes, are made using the Pim-1 protein. The tool combines docking, force field parametrization, and a free energy simulation and analysis into one overarching workflow. Both docking, parametrization, and the relative binding simulations of most of the metal complexes yield results that are accurate enough for further application/analysis. The accuracy of the parametrization method is found to be comparable to contemporary parametrization methods. Whereas the generated absolute binding affinities deviate significantly from literature values, the relative binding affinities largely correspond qualitatively and for some compounds quantitatively to literature. Our results exhibit the potential of the computational tool and highlight the need for further testing.