Shuming Liu

Hi, thank you so much for the reply! So this is the code I am using to serialize and check the type: ``` with open('%s/system.xml' % init_system_state_dir, 'w') as f:...

Ok, let me try. This may be slow as the system is large, it takes a long time to add forcefields.

So I am using this code to convert it to string: ``` with open('%s/system.xml' % init_system_state_dir, 'w') as f: system_xml = simtk.openmm.XmlSerializer.serialize(s) print('type(system_xml) = ', end='') print(type(system_xml)) str_system_xml = str(system_xml)...

I installed OpenMM by conda. I tried both openmm version 7.5.1 and 7.7.0 and they both show similar results. With the following code: ``` with open('%s/system.xml' % init_system_state_dir, 'w') as...

So may I ask if there is any restriction in OpenMM for system size, or restriction about the largest size of the system that can be output as .xml file?

So I did a further test, after building the system, I directly start the simulation with the system without saving it as xml. The simulation output looks fine. After finishing...

This is the environment for openmm 7.5.1 ``` _libgcc_mutex 0.1 conda_forge conda-forge _openmp_mutex 4.5 1_gnu conda-forge biopython 1.79 py36h8f6f2f9_0 conda-forge bzip2 1.0.8 h7f98852_4 conda-forge c-ares 1.18.1 h7f98852_0 conda-forge ca-certificates 2022.6.15...

So based on more tests, I think the issue is because the system is extremely large, not because the particle number is large, but because the forces are quite complex...

Hi Peter, sorry for the reply. So based on more testing, we think the issue is about the memory. Sorry it is hard to estimate the total file size, as...

I found by changing ``` mdtraj.rmsd(traj1, traj1[0]) ``` to ``` mdtraj.rmsd(traj1[:], traj1[0]) ``` then the coordinates will not be changed