Best Practices for Nucleobase Parameterization

From: Mayne, Christopher G (
Date: Wed May 20 2015 - 08:12:55 CDT


I see that you have also posted to the CHARMM user forums under the parameter discussions, which is what I was in the process of suggesting before I saw the post. They are a little touchy over any mention of VMD/NAMD over there; however, your question is very clearly on of the force field (not necessarily the specifics of the programs) and is appropriate for that forum. The CHARMM36 distribution has specific files that contain nucleic acid topology/parameters; the header of those files contains references that you might look over. I don't have any direct experience parameterizing NAs but I do have a few comments regarding your questions:

1) I'm fairly certain even the keto tautomer is planar, as the lone pair of the nitrogen is conjugated to the pi system of the adjacent carbonyl, just as is the case for any other amide.

2) I would be very surprised if the interactions with carbon and nitrogen of the base are not probed via water interactions to derive charges. Is this explicitly stated somewhere in the references for NA parameter development?

3) Cleaving the base from the sugar falls under the suggested divide-and-conquer strategy. I would be cautious of N1 and would cap it with at least a Me, which mimics the full NA structure moreso than cleavage to NH. As suggested to you over in the CHARMM forums, you will have to address the sugar-base linkage at some point as well.

Christopher Mayne

Date: Mon, 18 May 2015 12:30:22 -0400
From: Aaron Larsen <<>>
Subject: namd-l: Best Practices for Nucleobase Parameterization

Hi all,

Sorry if this message is a repeat, it appears to have not been sent
earlier. I am attempting to parameterize some RNA like nucleosides using
the wonderful Force Field Tool Kit (VMD 1.9.2) and I have a few questions
to ensure that my results are relevant:

1) The keto tautomers are likely to be of lower energy but the geometry of
nucleobase amino groups should, strictly speaking, be tetrahedral in this
form. This, of course, is not what is normally depicted for nucleobase
amino groups which are commonly depicted in planar geometry, representing
some character of the enol tautomer. In this case, is it best practice to
force the geometry to be planar for the amino groups?

2) When modelling the water interactions, convention seems to be to only
consider interactions with the aminos, carbonyls, methyls, and hydrogens
that decorate the nucleobase but not the interactions involving the carbons
and nitrogens of the rings themselves. Is this best practice?

3) To simplify the calculations, I am working on just the bases and not
including the sugars. This seems to be the convention with these sorts of
projects. Is there any concern that ignoring the sugars in this way will
result in bad parameters (especially for charge) for the N1 atom?

Thank you for your consideration.


- --
Aaron Larsen, Ph.D.
Harvard University Department of Chemistry and Chemical Biology
Harvard Medical School Department of Genetics
Mobile: 617-319-3782
FAX: 617-643-3328

This archive was generated by hypermail 2.1.6 : Thu Dec 31 2015 - 23:21:53 CST