Re: [AMBER] DMF

From: Thomas Cheatham <tec3.utah.edu>
Date: Wed, 2 Jan 2019 20:21:23 -0700 (MST)

> Enclosed file is a BOSS output file for OPLS parameters for DMF, I was
> wondering how can I use the information to generate an amber solvent
> parameters for DMF.

To my knowledge, there is no automated BOSS -> AMBER conversion tool,
however the underlying functional forms are similar. Consider this a good
(and necessary) learning experience to figure out how to generate the
appropriate files. This will help you understand the units used in each
program and hopefully amplify the critical need to test for yourself to
make sure DMF in AMBER matches DMF in BOSS. There are a number of recent
papers that compare how different programs do things, but nothing is
better than working through it yourself to trust the parameters,
conversion, and process... Trust involves running with the parameter files
and verifying that the calculated properties match (experiment or
previous simulation results with a different program).

Things to consider:

- making sure the units match
- electrostatic treatment (cutoff, Ewald, hard coded constants in code)
- combining rules for vdw (OPLS is geometric not Lorentz-Bertholet as in AMBER)
- time step
- pressure/temperature control
- constraints (SHAKE, Settle, ...)

Even if you have an automatic conversion program: trust, but verify, or
understand the risks... (Erratum?) Exactly reproducing results among
different programs / force field implementations is not trivial, however
it is possible to get (relatively) close agreement.

Reminds me to advise on MM-PBSA results; if they seem crazy, make sure you
can calculate by hand or alternative approach to verify they are crazy. If
crazy, try to understand why crazy and if they are actually real,
artifact, or error.

Recent paper on comparing ffs/programs:

Lessons learned from comparing molecular dynamics engines on the SAMPL5
dataset. Shirts MR, Klein C, Swails JM, Yin J, Gilson MK, Mobley DL, Case DA, Zhong
ED. J Comput Aided Mol Des. 2017 Jan;31(1):147-161. doi: 10.1007/s10822-016-9977-1.

--tec3

p.s. happy new year all!


_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Jan 02 2019 - 19:30:01 PST
Custom Search