Nowadays, the structures of individual proteins and many protein complexes can be accurately predicted using computational methods. Modern neural network-based methods also predict interactions between proteins and other molecules, including nucleic acids and organic ligands. However, these predictions are often less reliable than protein structure models. Moreover, obtaining a structure model for a biomolecular complex is usually not enough, and computational methods for further structure interpretation are way less accurate. As a result, in this project we propose to develop novel methods for the analysis of biomolecular interactions, based on the existing experimental data and also having the aim to better interpret the predictions from biochemical and biophysical point of view. The resulting structural bioinformatics tools are therefore expected to provide explainable and useful predictions for experimental researchers.
Mokslinis vadovas / Supervisor: Justas Dapkūnas
Kontaktai / Contacts:
Programme: Biochemistry N 004