From: Mcguire, Kelly (klmcguire_at_UCSD.EDU)
Date: Wed Aug 11 2021 - 01:09:13 CDT

I have setup a simulation that is about 2 microseconds long and ~1.5 million atoms. The protein is allowed to diffuse freely around the waterbox which includes Na and Cl ions. I am now doing analysis on the simulation. One of the things I need to do
is align the protein throughout the trajectory to a reference structure and then write out the frames as individual PDBs. When I align the protein to the backbone of the reference structure that works fine, but because the protein was free to diffuse around and tumble, the waterbox and ions rotate around the aligned protein. Does anyone have an idea on how to keep the waterbox and ions from rotating around the aligned protein? I need to convert the snapshots into densities and then be able to average the densitites of the whole system and if the waterbox is rotating around then the averaging will be messed up. The idea, really, is to have the protein in each of the aligned snapshots clustered together on the same reference but be able to see local movements in the sidechains and then make densities from those and average them. I hope that makes sense. Is there another way (instead of aligning to a reference structure) to undo the diffusion/tumbling and cluster the protein in each frame on the same reference while keeping the waterbox and ions from rotating around the protein? I suppose we could have put a strong center of mass constraint on the protein the keep it from diffusing around but allow local sidechain motion, but we need to include the full motion of the protein without constraints for what we are doing.

Dr. Kelly McGuire
Chemistry/Biochemistry Department
Natural Science Building, 4104A, 4106A, 4017