Protein benchmark simulations | Goal: figure out what the different tiers of tests are, prioritize. Figure out how to make sure analysis is using your scripts so that results are compatible. Us: Modify the scripts to allow OpenFF exports to GROMACS - validate with the first step to make sure energies are comparable. Tests: PDB structure, number of waters and ions. Documentation: I have most of that in a github Repo. How long should they be run? What do we run wirg? TODOS: CC - add more documentation to proteinbenchmark repo Lysozyme is done, Ubiquitin and BPTI. Anika to run GB3 (to verify). 10 us for folded proteins (defined in the parameters) How big should the water box should be, 12A bigger.
Run it through with OpenMM - configure the reporter in OpenMM - bond/angles/torsions so we can see in the initial file how those are from GROMACS. Then: Switch to GROMACS export and test. After generating 10 us or so, you want to see what you need to do to run the NMR scripts. IT MIGHT be something like mdtraj to load GROMACS .xtc files and write then out as OpenMM netcdf. OR: Using LOOS from Grossfield lab - can read in gromacs .xtc. Run on bridges to start. It worth it to figure out the scaling system are up to be, since we will be running.
|