VMD-L Mailing List

From: Fotis Baltoumas (fbaltoumas_at_biol.uoa.gr)
Date: Tue Sep 25 2018 - 02:01:04 CDT

Hello all,

After re-viewing my original reply and after an e-mail exchange, I would
like to amend an error in my first mail and explain a point about
electrostatics:

1. Contrary to what I originally wrote, ion CG particles are NOT four
ions grouped together but, rather, one ion grouped with three water
particles.  Should you decide to make a custom CGC file, you should take
that into consideration when building mapping CGC files.

2. The electrostatics in the Martini force field (and its derivatives,
RBCG, PACE etc) are a complex issue.  While it is true that there are
charged (Qd and Qa) particles, the fact that the force field a) uses a
uniform dielectric (15) for the entire system no matter what its
composition is and b) shifts electrostatics the way it does (0-12
Angstroms) means that electrostatics are often trivialized.  In theory
this shouldn't be a problem, as PMFs of model systems fit well with
all-atom ones. In practice, when electrostatics need to be seriously
considered, this has been shown to be an issue.

The Martini authors themselves have pointed this out on several
occasions, both in papers and in workshops.  A recent one, where it was
discussed, can be found  in the link below:

http://md.chem.rug.nl/images/stories/WORKSHOP2015/lecture-02.pdf

(see pages 10-18 for the details of parametrization, pages 37-41 for the
electrostatic problem)

Now in most situations electrostatics are not a problem, because they
mostly often covered by van der Waals interactions (the pair potentials
of the charged and polar beads have been parametrized to reproduce this
effect).  However, when a system requires careful screening of
electrostatics (as is the case with protein-ion complexes), this can
become a major problem.  Issues such as this one are the reason the
polarizable model was eventually developed.

In any case, sorry for the confusion my original answer may have caused.

Kind regards,
Fotis

On 09/25/2018 05:59 AM, Peter Freddolino wrote:
> Dear Kimberly,
> Just a couple of thoughts on this, which I'm posting partly because
> there have been a couple of questions about the VMD RBCG model lately.
>
> First, I'd like to point out that the RBCG model included in VMD is
> quite old at this point; one really ought to be using a recent martini
> version instead. As NAMD supports the martini FF now, all that would
> really be required would be to use appropriate cgc files and
> corresponding toppar files (see examples with the tutorial at
> https://www.ks.uiuc.edu/Training/Tutorials/martini/rbcg-tutorial.pdf)
> - one ought to double check that those are appropriate for the
> up-to-date version of the martini FF that you want to use. I raise
> this point only because the default RBCG conversion files used by vmd
> corresponds to a CG representation that really shouldn't be used for
> starting new projects any more since it is superseded by recent
> martini versions.
>
> Second, using recent martini versions will instantly fix your calcium
> problem (sort of), in that martini includes a topology for a
> (implicitly hydrated) calcium ion. Read the comments in the martini
> topology carefully for caveats.
>
> Third, you may want to think carefully, for your system, about whether
> it is better to treat your calcium ion as a calcium ion, or come up
> with some way of implicitly incorporating it into your protein
> topology. That might be a better option if it is a site bound calcium
> and you're not interested in letting it dissociate (especially as the
> martini free calcium is really supposed to represent calcium with a
> solvation shell).
>
> Best,
> Peter
>
>
>
> On Mon, Sep 24, 2018 at 2:27 PM, Lewis, Kimberly A
> <lewiskim13_at_students.ecu.edu <mailto:lewiskim13_at_students.ecu.edu>> wrote:
>
> Hello,
>
> When using VMD, specifically the residue based course grain
> builder, I know there is the protein and water sets that you can
> add to use. Is there something that will create Ions as a bead,
> that are bound to the protein such as Calcium? When I have tried
> so far, the beads are generated for the amino acids but the
> calcium ions are missing from the cg pdb.
>
>
> Thank You,
>
> Kimberly Lewis
>
>
> East Carolina University
>
> Department of Chemistry
>
> 

-- 
*******************************************
Fotis A. Baltoumas
Bioinformatics Postgraduate Programme
Section of Cell Biology and Biophysics
Department of Biology, National & Kapodistrian University of Athens
Panepistimiopolis, Athens 157 01, GREECE
   --------------------------------------
email : fbaltoumas_at_biol.uoa.gr
http://biophysics.biol.uoa.gr
http://bioinformatics.biol.uoa.gr
*******************************************