Srinivasaraghavan Kannan

• The application of Classical Molecular Dynamics (MD) for the structure prediction of biomolecules is limited by the accuracy of current force fields and the simulation time scale. Biomolecules can adopt several locally stable conformations separated by high energy barriers. Conformational transitions between these stable states can therefore be rare events even on the time scale of tens to hundreds of nanoseconds. Out of the various methods Replica Exchange Molecular Dynamics (Rex MD) is the most successful method to enhance the conformational sampling of biomolecules. But this is limited to only small systems, as the number of replicas required for Rex MD increases with increasing system size. Therefore, during my PhD, I have developed an alternative "Hamiltonian" replica-exchange method that focuses on the biomolecule backbone flexibility by employing a specific biasing potential to promote backbone transitions as a replica coordinate. The aim of this biasing potential is to reduce the energy barriers associated with peptide backbone dihedral transitions. The level of biasing gradually changes along the replicas such that frequent transitions are possible at high levels of biasing and thus the system can escape from getting trapped in local energy minima. This thesis discusses the development of this Biasing Potential Replica Exchange Molecular Dynamics (BP-Rex MD) method in detail. Application of the method to study the conformational sampling of peptides, folding of a mini protein and also for refinement and loop modeling of homology modeled proteins in explicit solvent shows much better sampling of conformational space as compared to the standard MD simulations. One of the main advantages of this BP-Rex MD simulation is that only the biasing potential energy term enters into the exchange probability, meaning that the number of required replicas is expected to scale approximately linearly with the number of included backbone dihedral angles.