GROMACS version: 2021.3
GROMACS modification: No
I have a .gro file with 2.7 million atoms from that I would like to create a top file in order to continue a simulation. However, using the following line on a single NVIDIA GeForce 2080 GPU, this job has not been completed after 2 days (which is hitting the job time limit for the cluster). Is there any way to more efficiently create a top file from a gro file? Thank you!
gmx_cuda pdb2gmx -f equil_1bar.gro -p equil_1bar.top
There is no GPU acceleration for pdb2gmx so it’s only going to run on whatever CPU resources you provided for it. I suspect a lot of those 2.7 million atoms are redundant things like solvent, lipids, etc. which means you should really strip the system down to simpler components, run those through pdb2gmx, and add in the repetitive species as simple #include statements so you don’t have to explicitly generate a topology for a million or more solvent particles. It’s just unnecessary work.