Re: query about the cross term in psf

From: Josh Vermaas (vermaas2_at_illinois.edu)
Date: Sat Dec 06 2014 - 14:22:32 CST

The advice doesn't really change. You shouldn't be deleting terms willy
nilly. Crossterms = CMAP correction, and without those crossterms, you
fall back (I think) to something like charmm22, the last force field
release before CMAP was added. Figure out where in your psf-building
process that crossterm is coming from, because something weird is going
on. NR3 is an atomtype off on the sidechain of HIS, so it should not be
involved in CMAP, which is there to get the backbone phi/psi right.
-Josh

On 12/06/2014 02:12 PM, Kshatresh Dutta Dubey wrote:
> Oh sorry Josh, perhaps I pasted wrong error sentence, initially it was
> dihedral terms but I added those from previous lipid paramters and
> rectified in parameter file. But after that when I re-ran the job it
> was giving error:
> FATAL ERROR: UNABLE TO FIND CROSSTERM PARAMETERS FOR C NH1 CT1 C
> NH1 CT1 C NR3.
> Therefore please suggest me about the cross-term.
> Thanks
>
> On Sat, Dec 6, 2014 at 9:49 PM, Josh Vermaas <vermaas2_at_illinois.edu
> <mailto:vermaas2_at_illinois.edu>> wrote:
>
> Cross terms or dihedrals? The error is quite clear. You gave NAMD
> a psf that had a dihedral (*not* a crossterm) term it did not
> recognize, and to prevent you from doing something you didn't
> intend, it terminated for you to avoid wasting computer time.
> Where did that term come from? Grepping around in the charmm
> topology files, NR3 only shows up in proteins, and HCL is an
> atomtype seen in lipids. Deleting terms wholesale from a psf is a
> downright *dangerous* thing to do, as you will no longer be
> simulating with a force field that has in any way been validated.
> Is there something special about the system that needed a linkage
> between proteins and lipids? If not, do you have any idea how that
> dihedral term got into your psf?
> -Josh Vermaas
>
>
> On 12/06/2014 01:31 PM, Kshatresh Dutta Dubey wrote:
>> Dear Users,
>>
>> I prepared the psf and corresponding pdb using vmd, but it gave
>> error :
>> UNABLE TO FIND DIHEDRAL PARAMETERS FOR O C NR3 HCL (ATOMS 7422
>> 7421 7427 7448)
>>
>> I googled the error and found somewhere that deleting the
>> cross-terms in psf can make job to run. I did and job is running
>> now. However, I am curious to know whether to delete cross-terms
>> in psf is correct or not. Please suggest me.
>> Thanks in advance.
>>
>> --
>> With best regards
>> ************************************************************************************************
>> Kshatresh Dutta Dubey
>> Post Doctoral Researcher,
>> Lise Meitner Center for Computational Quantum Chemistry
>> Hebrew University of Jerusalem Israel
>> Jerusalem, Israel
>>
>>
>
>
>
>
> --
> With best regards
> ************************************************************************************************
> Kshatresh Dutta Dubey
> Post Doctoral Researcher,
> Lise Meitner Center for Computational Quantum Chemistry
> Hebrew University of Jerusalem Israel
> Jerusalem, Israel
>
>

This archive was generated by hypermail 2.1.6 : Wed Dec 31 2014 - 23:23:06 CST