CCL: Release of R.E.D. Server 2.0
- From: "Venable, Richard (NIH/NHLBI) [E]"
- Subject: CCL: Release of R.E.D. Server 2.0
- Date: Fri, 5 Nov 2010 15:12:55 -0400
Sent to CCL by: "Venable, Richard (NIH/NHLBI) [E]"
I've read through the paper and the two specific CHARMM citations, and have to
admit my first response over-assumed a reliance on RESP charges. However, I am
still concerned about the claim to be to produce complete force field libraries
for CHARMM, which I initially interpreted as ready-to-use residues in the proper
topology file format, which wasn't the case. I believe a claim of being able to
produce entire CHARMM compatible force field libraries should be demonstrated
much more completely.
This quote from the cited paper itself raises question about the level of CHARMM
and OPLS compatibility:
"Indeed, RESP and ESP charge values derived using R.E.D. are fully
compatible with Amber and GLYCAM force fields, (40-47) and can be used in CHARMM
and OPLS force field based simulations as well. (17-20)"
References 19 and 20 are not really demonstrations of the exact use of R.E.D.,
but use CHELPG for the initial charges, which are further adjusted for CHARMM
Reading further, it seems the burden of choosing the correct LJ types, and
thereby the mapping to angle and dihedral terms, is left completely up to the
end user, which I see as a major weakness. Thus, the type independence may pose
a problem, in that novel bonding geometries can be created for which there are
no angle and dihedral parameters during the user mediated atom typing process.
Rick Venable 5635 FL/T906
Membrane Biophysics Section
NIH/NHLBI Lab. of Computational Biology
Bethesda, MD 20892-9314 U.S.A.
(301) 496-1905 venabler AT nhlbi*nih*gov
On 11/4/10 10:09 PM, "Barry Hardy barry.hardy*o*vtxmail.ch"
Sent to CCL by: FyD [fyd : q4md-forcefieldtools.org]
Dear Rick Venable,
R.E.D. means RESP and ESP charge Derivation.
The R.E.D. Tools and R.E.D. Server do not only deal with _RESP_
charges but with _ESP_ charges using various algorithms used in MEP
computation. And ESP charges are used in FFs different from AMBER FFs...
Then, charge derivation is the first step - charge validation is done
in a second step.
See http://www.ncbi.nlm.nih.gov/pubmed/20574571 & references
This is why we wrote
"R.E.D. Server is a web service designed to automatically derive RESP
and ESP charges, and to build force field libraries for new
molecules/molecular fragments. R.E.D. Server provides to computational
biologists involved in AMBER, CHARMM, GLYCAM & OPLS force field based
biological studies the software and hardware required for charge
derivation and force field library building."
> Sent to CCL by: "Venable, Richard (NIH/NHLBI) [E]"
> Given that RESP/ESP methods are **not** used for the development of
> CHARMM force fields (a fragment approach similar to OPLS is used),
> the compatibility of molecules with RESP-based charges with the rest
> of the CHARMM force fields is somewhat questionable. The use of
> other approaches, notably the CHARMM General force field (CGenFF),
> is recommended instead of ad hoc web servers for extending the
> CHARMM force field to new molecules.
> I did not see anything on the q4md site which discussed this, or
> that gave any hints about validation of the molecular descriptions
> produced in the context of CHARMM and its distributed force fields.