Vejleder: Himanshu Khandelia
Projektbeskrivelse:
The goal of this project is to develop and validate a computational method that generates, refines and ranks linear or cyclic peptide binders against a chosen protein target by combining machine-learning generation with physics-based simulations.
Methods such as RFDiffusion and Boltz can design peptides to bind protein targets. In our experience, however, the performance of these methods is limited owing to the lack of physicochemical validation. We will combine physics-based molecular dynamics simulations, machine learning, and biochemical experiments to develop new peptides targeting specific enzymes involved in RT-qPCR tests. The newly designed peptides are likely to significantly reduce the cost of RT-qPCR tests.
OBS!: The students are expected to have a working knowledge of Python.