GNEIMO Constrained Dynamics Method: A Tool for Protein Structure Refinement and Conformational Changes

Constrained molecular dynamics methods, wherein the high frequency degrees of freedom are placed as hard holonomic constraints in the dynamic model of the protein have been developed several decades ago, but these methods are not used widely. There are several bottlenecks in using these methods, the most formidable being the computational time taken for solving the coupled equations of motion. Spatial operator algebra(SOA) techniques reduces the computational time for solving the equations of motion by two orders of magnitude. We have developed a computational framework called GNEIMO, that uses the SOA techniques combined with all-atom force field and appropriate integrators to solve the constrained equations of motion. The generalized constrained molecular dynamics method GNEIMO, allows the user to “freeze and thaw” torsional degrees of freedom as fit for the problem studied.We will demonstrate the use of GNEIMO method in protein structure refinement of low resolution homology models. Starting from low resolution homology models we observed that the all-torsion GNEIMO method leads to a 2A improvement in RMSD to the crystal structure, while the all-atom molecular dynamics method disrupts the starting model further. The GNEIMO method also showed enrichment in the population density of native-like conformations.We have also tested out the GNEIMO method for studying conformational transitions between two well characterized (crystal structures) conformations of a protein. Long time scale dynamics with GNEIMO on calmodulin and fasciculin shows that the transitions from one conformation to another happen with more facility than with all-atom MD simulations.