I recently found that when running langevin dynamics in charmm with openmm, the seed saved in rst file was not changing (while the rst file was updating). When I tested it with charmm alone without openmm, there is no problem and the seed in rst was updating correctly.
I think and hope this might be a small problem: the random sequence was actually generated by openmm and dynamics was carried out correctly. It is just for some reason charmm failed to get the information from openmm.
If this is the case, I can provide "iseed" everytime I run "dynamics" command. With inside the "dynamics", the seed assigned by "iseed" will generate the random sequence.
However, there is a chance that something was wrong and the random sequence was not generating at all inside openmm. The same number was used again and again with inside the "dynamics" command, and it was passed back to charmm correctly so the number did not change in rst file.
So how can I identify which is the case? Is there anything in the output can give me some clue about it?
Thank you very much.