Skip to content

Trusted force field files for gromacs

Notifications You must be signed in to change notification settings

moleculadesigner/gromacs_ff

 
 

Repository files navigation

gromacs_ff

Trusted force field files for gromacs in this repo:

General notes on available force fields

Simulating DNA

  • The best source of information is usually the Amber manual. Currently, Amber 19 manual on page 38 discusses two options parmbsc1 from Orozco group and OL15 from Sponer group. The do not make a choice as to which is better, howerver, on the front page the OL15 is recommended.
  • As for ions, by default the Young Cheatham 2008 ions are selected in Amber ff. Looks like it may be a good idea to replace them with CUFIX from Aksimentiev group.
  • The amber14sb_parmbsc1_cufix is available in this repo.
  • The amber OL15 can be downloaded from the gromacs web-site.

Simulating disordered proteins

  • There is CHARMM36m (2016) ff, it is now recommended for all protein simulations, can be found at MacKerells web-page (see below). Also ports for Gromacs are there.
  • In AMBER another way is to use OPC water model (see this paper https://pubs.acs.org/doi/10.1021/acs.jctc.8b01123) with 12-6 HFE Li and Merz ion parameters. We need to add a gromacs prot here ...

Simulations of proteins and protein folding

  • Generally there are two options standard latest AMBER or CHARMM force fields.
  • See AMBER manual for AMBER, currently the ff14SB is recommended.
  • See MacKerell web-site for CHARMM http://mackerell.umaryland.edu/charmm_ff.shtml . Strictly speaking the charmm ff is updated with each new release of charmm (currently 44). However, since CHARMM36 the changes have been small, so on MacKerell's web-page they are reffered as updates to the CHARMM36 ff.

Links to external ff resources

AMBER

About

Trusted force field files for gromacs

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Rich Text Format 98.8%
  • Makefile 1.2%