Pure POPC membrane simulations with 1000 mM CaCl2 with the CHARMM-Drude force field (OpenMM)

400 ns MD simulation of pure POPC membrane using Charmm-Drude polarizable force field. The system contains 128 POPC lipids, 114 CaCl2, and 6400 SWM4 water molecules.

Before running the Drude simulation, the system has been equilibriated using Charmm36 force field for 200 ns. The last frame of that simulation was used to generate Drude polarizable model. The first 100 ns of the Drude simulation has been discarded from this dataset. Total simulation time is 500 ns, included data is 397.5 ns.

It has been discovered that (https://github.com/NMRLipids/Databank/issues/2#issuecomment-1357871243) the previously uploaded trajectory did not have the correct timestamp: the timestep between two consecutive simulation frames was not correctly embedded into the trajectory information. Therefore, with the latest version we are uploading the "wrapped_full.xtc" which has the correct timestamp. The frame saving frequency in this trajectory is 10 ps and there are 39750 frames.

This new update should not invalidate any previous calculations that did not explicitly read the timestamp information from the trajectory.

This simulation consists of 4 sub-trajectories, each of which starts from the last frame of the previous one. These trajectories (originally in dcd format) were concatenated and saved in xtc format with MDAnalysis.

Centering of the trajectories has been done via below MDAnalysis script

    ...:     u = mda.Universe('../step3_charmm2omm.psf', 'step5.dcd')
    ...:     prot = u.select_atoms("resname POPC")
    ...:     ag = u.atoms
    ...:     workflow = (transformations.unwrap(ag),
    ...:     transformations.center_in_box(prot, center='mass'),
    ...:     transformations.wrap(ag, compound='fragments'))
    ...:     u.trajectory.add_transformations(*workflow)