Page 1 of 1

Nucleic acid Tinker 2018 parameters to xml conversion

Posted: Sat May 18, 2019 6:23 am
by mandar5335
Hi all,
I am a new user to openmm as well as tinker-openmm softwares.

I wish to use recently published (DOI: 10.1021/acs.jctc.7b01169) nucleic acid AMOEBA parameters using only openmm-plumed combination. But I am unable to find way for conversion of new tinker parm file to xml format.

some of the old discussions on forum suggest use of parmED, but the follow up discussion showed it was not possible.

Any suggestions or comments will be really helpful.

Thanks in advance,
Mandar Kulkarni
postdoctoral fellow,
CMPS, Lund University

Re: Nucleic acid Tinker 2018 parameters to xml conversion

Posted: Mon May 20, 2019 4:15 pm
by peastman
We haven't yet converted the nucleotide parameters to OpenMM format. Thanks for pointing that out. We really ought to.

The conversion of AMOEBA was originally done with the script at https://github.com/pandegroup/openmm/bl ... ceField.py. In principle it should be easy to use that script to convert the new parameters, though I suspect some changes will be needed to get it to work. If you'd like to try, that would be great. Otherwise I'll put it on my to-do list, though I can't promise when I'll get to it.

Re: Nucleic acid Tinker 2018 parameters to xml conversion

Posted: Tue May 21, 2019 12:37 am
by mandar5335
Thanks for your reply. I would like to modify the suggested script. I saw this script before but I am not sure how to use it.

Please is it possible to provide some suggestions on correct usage of this script? Also, if possible, could you please point out the suspected changes that I should keep in mind while modifying the script.

Re: Nucleic acid Tinker 2018 parameters to xml conversion

Posted: Wed May 22, 2019 12:08 pm
by peastman
Just run the script, specifying the file to convert as an argument:

Code: Select all

python processTinkerForceField.py amoebabio18.prm
Once you do that, hopefully everything will just magically work and you'll get a correct force field as output. But I also won't be at all surprised if that doesn't happen. Quite likely it will either 1) throw an exception, or 2) produce an incorrect force field. Any exceptions will hopefully be easy to fix. To ensure correctness, we'll need to compute forces and energies for a bunch of systems and compare them to values computed with Tinker.

Re: Nucleic acid Tinker 2018 parameters to xml conversion

Posted: Fri May 24, 2019 2:17 am
by mandar5335
Thanks again for the suggestion.
I have run the script and it works as you suggested. But, I think script is missing conversion of parameters associated with phosphate groups mainly "strtor" and "angtors". I am unable to find the conversion of these terms to openmm format.

If you can point me to some manual or section describing this conversion, I will try to add these modules.

The following URL link provides log file in case you want to see if I am missing anything else.

https://www.dropbox.com/sh/vn8vsf8xk4gl ... BD1Ka?dl=0

Re: Nucleic acid Tinker 2018 parameters to xml conversion

Posted: Tue May 28, 2019 10:03 am
by mandar5335
I realized that these Tinker's stretch-torsion and angle-torsion coupling forces are not present in openmm and thus, it's not possible to use OpenMM-Plumed combination right now for nucleic acid simulations.

I am just curious is there any near future possibility that these forces will be included, thus allowing OpenMM-Plumed combination eventually?

Thanks again for all your suggestions.

Re: Nucleic acid Tinker 2018 parameters to xml conversion

Posted: Thu May 30, 2019 1:32 pm
by peastman
It's actually easy enough to implement them using a CustomCompoundBondForce. The problem is that ForceField doesn't have any built in mechanism for creating them. There are workarounds for that too. For example, the XML file could include an embedded Python script that constructs them. That's distinctly more complicated though.