Re: NAMD and usable FFs

From: Francesco Pietra (
Date: Thu Oct 16 2014 - 16:41:59 CDT

Again to CGenFF after a long while, I was terrorized by the complexity of
symbols for atom types of organic molecules (protein ligands). Also, for
ligands not too far away from purine bases, CGenFF performed quite naively,
for example, giving a single Kb for bonds, while antechamber gave correctly
much higher Kb for C=O than C-C. So, I built bonds, angles, dih, impr from
antechamber. Then, apart from the problem of charmm-type partial charges, I
know how to move from antechamber to charmm with the obsolete all27 and
all22, I have to think about using all36 here. I must admit that I never
spent adequate time in learning how to parameterize organic ligands by
analogy in the charmm world. I tried a couple of times ffTK just to be on
an absolute basis, a very nice suite but extremely time consuming. Nothing
good comes for little, however. At any event, moving from the old
organization of charmm ff to all36 is quite convincing from your reasoning,
that getting no MD crash is no assurance per se to be on the right way.
Straightforward with normal proteins, the problems come with organic
ligands, not to say with metalloproteins. In that respect, which is nearly
the norm for me, amber offers easier tools, although renumbering residues
makes life difficult in case of multimers.
Thanks a lot to you and Josh.

On Thu, Oct 16, 2014 at 8:45 PM, Kenno Vanommeslaeghe <> wrote:

> True, grep totally cannot be relied upon for this purpose, to the extent
> that it's a recipe for disaster. But rather than going in and messing with
> the files manually (the mileage of which may vary depending on user skill),
> why not just load everything into psfgen / NAMD ? The programs can handle
> it, and with CHARMM36, we essentially eliminated the potential for
> collisions.
> On 10/16/2014 02:05 PM, Josh Vermaas wrote:
>> Hi Francesco,
>> I use psfgen built into VMD, so any legal tcl commands would work. I think
>> glob and grep commands in tcl can be made to do what you want, but I'd
>> strongly advocate doing the grepping manually in a terminal, and
>> inspecting the residues that pop up to make sure you know what you are
>> getting (protonation states, missing linkers, etc). Spending an extra
>> (half?) hour of your time to look through the tree yourself will be time
>> well spent if it saves you from needing to redo all your simulations
>> because of a "clever" grep picked up something you didn't intend or is
>> missing some important terms.
>> -Josh
>> On 10/16/2014 12:59 PM, Francesco Pietra wrote:
>>> Hi Josh:
>>> Thanks for guidance.
>>> Would it be possible to combine the "grep <myname> *" command into
>>> psfgen? Just to avoid to extract the needed particular topology from
>>> toppar_c36.
>>> Thanks again
>>> francesco
>>> On Wed, Oct 15, 2014 at 4:20 PM, Josh Vermaas <
>>> <>> wrote:
>>> Hi Francesco,
>>> I believe they all are, although some of the combined toppar files
>>> used to cause NAMD problems in 2.9 (not sure about 2.10) with
>>> unrecognized commands. The parameter components themselves are all
>>> perfectly acceptable, so I would split the distributed toppar files
>>> up into separate components (top and par) and everything would work
>>> out just fine. For your specific question, its probably safe to
>>> switch to using CHARMM36 for the protein components, although it'd
>>> be best for you to convince yourself of that by reading the more
>>> recent papers describing the development of the CHARMM protein
>>> forcefield.
>>> -Josh Vermaas
>>> On 10/15/2014 04:34 AM, Francesco Pietra wrote:
>>>> Hello:
>>>> It is not clear to the general - mostly experimental - user, which
>>>> CHARMM FFs for proteins, metalloproteins, and organic ligands are
>>>> compatible with NAMD 2.10.
>>>> For example, with namd2.10 I currently use
>>>> toppar_all22_prot_heme.str for dioxygen in combination with
>>>> top_all27 and par_all27 for the protein. Could I use a more recent
>>>> version of CHARMM FF with that toppar? This is just an example, a
>>>> general table of compatibility would be most welcome, especially in
>>>> view of using CGenFF.
>>>> thanks
>>>> francesco pietra

This archive was generated by hypermail 2.1.6 : Wed Dec 31 2014 - 23:22:56 CST