Hello, A change in allocation has been requested: Requester: roux (Benoit Roux) Project: Drude Title: Development and validation of new polarizable force field for molecular dynamics Description: We are mainly interested in ions binding to proteins at this point. In biomolecular simulations, the challenge is to construct functional forms that are amenable to computationally efficient simulations while remaining able to accurately capture the microscopic interactions between molecules. Our efforts, done in collaboration with Alex MacKerell, have been to develop a polarizable force field for proteins, membranes, and nucleic acids based on classical Drude oscillators. The Drude polarizable force field has been implemented in the program NAMD. It runs efficiently, at a computational cost comparable to that of non-polarizable force fields (slower only by a factor of about 1.4).By now, the parameterization for water, ions, phospholipids, and proteins has been done and we are entering an exciting phase where we will start to apply the new force field to real systems and investigate a wide range of processes (ion selectivity, ligand binding, pKa, membranes, etc ). As an example, we have recently started to directly tackle the conditions under which selectivity has been quantitatively measured for KcsA, namely from the Ba2+ block experiments of Piasta and Miller.66 Now we have an accurate polarizable force field, including divalent ions,67 proteins and lipids,68 and we are in a position to carry out these calculations on the KcsA channel with its activation gate open. Comparison of the Drude model with density functional theory (DFT) calculations at the B3PW91/lanl2dz level shows that one Ba2+ coordinated by two N-methylacetamides have an energy of -133.0 kcal/mol compared to -131.4 kcal/mol with the Drude model. Preliminary MD simulations of KcsA blocked by Ba2+ with the Drude force field for the protein, lipids, water, and ions show that the system remains close to the X-ray configuration. We want to optimize ion parameters for the KcsA potassium channel and then simulate the blockade by barium ions. For this we will have to run classical MD simulations (with a force field), extract snapshots of the ion with its surrounding atoms, and then use compute energies and forces from ab initio methods as target data to optimize the force field parameters. Most of the classical simulations will be executed elsewhere, but the ab initio calculations will be carried out with Gaussian on fusion. We also want to run QM/MM simulations of the same set of atoms using CPMD to compare the atomic forces from force field and QM. We expect that about 200 snapshots for each ion type should do the job. We have to consider K+, Na+ and Ba2+. There are about 250 atoms to consider in the QM region. Current: undetermined amount Justification: Gaussian and CPMD are established programs that runs efficiently on fusion. These calculations are fairly standard. Requested: 875000 A specific reason has been given: These SUs are urgently needed to complete the LDRD project on force field optimization before the end of FY1. This needs to be approved and the final allocation amount decided upon. Thank You, The LCRC Accounts System