search

mosdef-hub / gmso

Flexible storage of chemical topology for molecular simulation
https://gmso.mosdef.org
MIT License
52 stars 33 forks source link

File format support #6

Open mattwthompson opened 5 years ago

mattwthompson commented 5 years ago

I don't think GitHub has the concept of a pinned issue, but I want to have a master tracker of the status of support for various file formats.

Extension Engine Typed or Un-typed? Internal reader Internal writer External support
.xyz many Un-typed ✅ 8bc93601e8eb26f7eb9586b05bd16920e8977a0d ✅ #46
.mol2 many Either
.gro GROMACS Un-typed ✅#26 ✅ #48
.top GROMACS Typed
.pdb many Un-typed
.data LAMMPS Typed ✅ #28
.gsd HOOMD Typed ✅ #47
? Cassandra Both?
.par GOMC/NAMD? #78 Typed
.psf GOMC/NAMD #78 Un-typed
.zmat ilff/fftool Un-typed
mattwthompson commented 5 years ago

Not sure if this should be in topology.py or elsewhere, but we eventually want to wrap each of these functions into a single, simpler one, like mb.load() or pmd.load_file(), that infers file type based on extension and calls the appropriate reader.

mattwthompson commented 5 years ago

We should think about how best to test our writers. Obviously we'll want to make sure the files get written, but it's not so easy to verify they get written accurately. Andrew brought this up in mbuild years ago (https://github.com/mosdef-hub/mbuild/issues/217) and basically no progress has been made because it's hard to do. The only way I can think of breaking this up is

  1. Make sure our readers can read the files back in
  2. Make sure other readers can read the files back in (picking them on a case-by-case basis)
  3. Sanity checks and agreement between the above
  4. Evaluate the total potential of the system (much farther down the road)
mattwthompson commented 5 years ago

UMN's CP2K repo https://github.com/ramanishsingh/CSSI-CP2K