From: JC Gumbart (gumbart_at_physics.gatech.edu)
Date: Mon May 23 2016 - 08:59:24 CDT
Whatever angle psfgen sets the H at is determined by (1) internal coordinates in the topology file (the IC lines) and (2) some simple heuristics. That angle should not be considered final, but just a first guess; run a minimization to see what the resulting preferred angle is.
If I understand you correctly, you’re talking about fixing the angles to what they should be in simulation, but all that does is push the inaccuracy in the force field somewhere else. I don’t think you can claim it gives you any particular insight.
I have no idea about future psfgen support for Drude - we use charmm tools (including charmmgui.org) to prepare files for those simulations. But I suspect Drude (or another polarizable force field) is the only possible solution for what you want. Or you have to ask less precise questions about the system.
Best,
JC
> On May 22, 2016, at 2:46 AM, Francesco Pietra <chiendarret_at_gmail.com> wrote:
>
> Sorry, replace "in my case 11 instead of 180" with "in my case 111 instead of 180"
> ---------- Forwarded message ----------
> From: Francesco Pietra <chiendarret_at_gmail.com <mailto:chiendarret_at_gmail.com>>
> Date: Sun, May 22, 2016 at 8:43 AM
> Subject: Re: namd-l: H-bonds in charmm36 with namd2.11
> To: JC Gumbart <gumbart_at_physics.gatech.edu <mailto:gumbart_at_physics.gatech.edu>>
> Cc: NAMD <namd-l_at_ks.uiuc.edu <mailto:namd-l_at_ks.uiuc.edu>>
>
>
> What about setting distance and angle colvars (with namd2.11/charmm36) about key H-bonds, as determined by H-ray diffraction, and built on psfgen (psfgen fails to set the angle correctly, in my case 11 instead of 180 degrees)? In order to study what happens elsewhere during MD.
>
> The problem is most acute with metalloproteins, with the active center built around d-block transition metals. Which is my case. Currently hopeless for me using polarized ff, although my model for the metal ion is polarized, though not in the frame of available polarized ff.
>
> Incidentally, is any future prospect for psfgen with Drude or other polarized potential?
>
> Thanks
>
> francesco pietra
>
> On Sun, May 22, 2016 at 1:44 AM, JC Gumbart <gumbart_at_physics.gatech.edu <mailto:gumbart_at_physics.gatech.edu>> wrote:
> There’s a few factors at play here, the main one being lack of polarization. Beta sheets in particular can not reproduce the correct geometry of hydrogen bonds in a fixed-charge force field (at least that’s what I recall).
>
> Here’s a recent paper that seems especially relevant: http://pubs.acs.org/doi/abs/10.1021/jp503972j <http://pubs.acs.org/doi/abs/10.1021/jp503972j>
>
> Best,
> JC
>
>
>> On May 20, 2016, at 5:18 PM, Francesco Pietra <chiendarret_at_gmail.com <mailto:chiendarret_at_gmail.com>> wrote:
>>
>> Hello:
>>
>>
>>
>> H-bonds in the charmm-ff are substantially viewed as in the excerpts below:
>>
>>
>>
>> From: Tim Isgro (timisgro_at_ks.uiuc.edu <mailto:timisgro_at_ks.uiuc.edu?Subject=Re:%20%20hydrogen%20bonds%20in%20NAMD>)
>> Date: Fri Mar 19 2004 - 11:25:47 CST
>>
>> Next message: Streiff, John H.: "initial temp of simulation" <http://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l.2003-2004/0720.html>
>> Previous message: Stern, Julie: "running multiple minimizations" <http://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l.2003-2004/0718.html>
>> In reply to: Ioana Cozmuta: "hydrogen bonds in NAMD" <http://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l.2003-2004/0710.html>
>> Messages sorted by: [ date ] <http://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l.2003-2004/date.html#719> [ thread ] <http://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l.2003-2004/index.html#719> [ subject ] <http://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l.2003-2004/subject.html#719> [ author ] <http://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l.2003-2004/author.html#719> [ attachment ] <http://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l.2003-2004/attachment.html>
>> Hi Ioana,
>>
>> The hydrogen bond potential in the Charmm force field is, from what I
>> gather, viewed skeptically by many people, since there is no very good way
>> to benchmark the parameters and potential.
>> From: Axel Kohlmeyer (akohlmey_at_gmail.com <mailto:akohlmey_at_gmail.com?Subject=Re:%20%20question%20on%20hydrogen%20bond>)
>> Date: Mon Sep 06 2010 - 04:08:32 CDT
>>
>> Next message: Ramya Narasimhan: "user-defined energy function" <http://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l.2009-2010/3673.html>
>> Previous message: dhacademic: "Re: FATAL ERROR: Due to a design error, GlobalMasterServer does not support individual atom requests from multiple global force clients on parallel runs." <http://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l.2009-2010/3671.html>
>> In reply to: Kwee Hong: "question on hydrogen bond" <http://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l.2009-2010/3670.html>
>> Next in thread: ehenriques_at_qui.uc.pt: "Re: question on hydrogen bond" <http://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l.2009-2010/3674.html>
>> Messages sorted by: [ date ] <http://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l.2009-2010/date.html#3672> [ thread ] <http://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l.2009-2010/index.html#3672> [ subject ] <http://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l.2009-2010/subject.html#3672> [ author ] <http://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l.2009-2010/author.html#3672> [ attachment ] <http://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l.2009-2010/attachment.html>
>> On Sun, Sep 5, 2010 at 11:40 PM, Kwee Hong <joyssstan0202_at_gmail.com <mailto:joyssstan0202_at_gmail.com?Subject=Re:%20%20question%20on%20hydrogen%20bond>> wrote:
>> > Hi.
>>
>> joyce,
>>
>> > In the namd-tutorial-unix.pdf, it mentioned that explicit hydrogen bond
>> > terms are no longer present in the CHARMM force field and are therefore not
>> > calculated by NAMD. Does this mean that throughout the simulation, hydrogen
>> > bonding are not calculated? So does that mean that I can't inspect the
>>
>> hydrogen bonds are a phenomenological description and thus
>> technically a not an easily identifiable term for a force field.
>> they are more a "side effect" of the charge distribution and
>> geometry of groups involving hydrogens bound to certain atoms
>> (h-bond donors) and their counter parts (h-bond acceptors
>>
>>
>>
>> Be that as it may be, is charmm36, as run with namd2.11 for a protein in TIP3 water solution, capable of maintaining all distances between aa O-atoms with interposed H atom, as determined by X-ray diffraction of the protein crystal? When not (as I am observing), should one suspect that the fictitious charge distribution was incorrectly calculated? Or simply assume that crystal and water solution are different affairs? Which would imply that crystallographers should refrain from making too many conclusions from their structural data.
>>
>> Grateful for a recent point of view.
>>
>> francesco pietra
>
>
>
This archive was generated by hypermail 2.1.6 : Tue Dec 27 2016 - 23:22:11 CST