This Master’s thesis describes two novel protocols developed to lower the cost of Markov state model construction for the glycine receptor: i) transition path sampling; and ii) the perturbative approach. Coarse-grained transition path sampling with eBDIMS, combined with atomistic detail reconstruction using cg2all reproduced biologically relevant low-energy intermediate structures between preexisting end-state structures. On the other hand, perturbative ligand morphing and ligand insertion were utilized for the creation of new structures based on knowledge about related systems. Maximum common substructure alignment (MCS) proved useful for ligand morphing, while a primitive alignment approach outperformed docking for the insertion of ligands. The newly developed methods were applied to the agonist-bound glycine receptor. Notably, perturbative ligand insertion showed promise in significantly reducing the computational cost of sampling the entire conformational ensemble that is needed for the subsequent creation of a Markov state model of the biological system.
|
