[gmx-users] Ion binding to DNA

[Justin Lemkul]

On 6/11/15 1:40 AM, soumadwip ghosh wrote:
> Hello,
>           I am asked by one of the reviewers of my paper about the
> distribution of ions ( Mg2+) bound to a single stranded DNA as a function
> of simulation time at different salt concentrations. In order to do that I
> have calculated the RDF of ions from the closest atom of DNA using g_rdf
> with surf mol and res_com option at different time intervals. The first
> peak minima came around 5 angstrom and I calculated the number of ions in
> that cut-off at different time intervals with the -cn option of g_rdf. Then
> I am plotting it as no.of ions bound to DNA against time. what I want to
> ask is there anything wrong with this approach? I have referred to the work
> below
>
> http://www.tandfonline.com/doi/pdf/10.1080/07391102.2012.732344
>
> Is there some other ways ( command or tool wise ) where the calculations
> are more accurate as far as the ion binding approach is concerned?
>

Dealing with Mg2+ is complex.  The above paper draws similar to conclusions with 
what others have seen.  Additive models of divalent cations are generally pretty 
crude.  Mg2+ has a very slow (us - ms) exchange rate with water and about a 12.5 
kcal/mol barrier for water dissociation.  You're never going to see exchange in 
an MD simulation, so if there are direct interactions with the nucleic acid 
(which is extremely important for RNA, but I've never heard of that for DNA) 
then those won't happen unless you place the ions there specifically or take 
them from a crystal structure.  The slow dissociation is part artifact, part 
reality.  Tricky proposition.

As for the issue at hand, I think your method of RDF is very crude.  It is 
generally of much greater interest to identify the functional groups that 
coordinate the ions.  You are unlikely to have direct Mg2+ interactions given 
the above information, but proximity to actual groups (phosphate, base moieties) 
is interesting information that you're completely glossing over.

-Justin

-- 
==================================================

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul

==================================================