In this proposed project, we plan to use different cutting-edge simulation techniques to unravel the role of antibody binding site rearrangements in antigen binding and investigate the influence of different heavy and light chain pairings, mutations in the framework and in the CDR loops on the specificity and physicochemical properties.
We will use available crystal structures and state-of-the-art antibody structure models as starting structures for our simulations, as we are confident that our sampling can capture long time-scale rearrangements and functional conformational changes of the antigen binding site.
Therefore, we will apply well-established enhanced sampling techniques in combination with classical molecular dynamics simulations to obtain conformational ensembles in solution and to reconstruct kinetics and thermodynamics of different paratope states in solution with Markov-state models.
Current limitations in antibody structure prediction and design can be overcome by characterizing the antigen binding site as interconverting paratope states, which play a critical role in antigen recognition.
As the antigen-binding fragment contains approximately 500 residues, the required simulation time to capture essential conformational rearrangements of the binding site is enormous. Additionally, the programs we used are GPU compatible and the methods we applied are easily parallelizable, making this project suitable for large scale supercomputers.