From: Jim Phillips (jim_at_ks.uiuc.edu)
Date: Tue Feb 17 2015 - 13:53:04 CST
Hi Brian,
This is RESI ASP from the topology file:
BOND CB CA CG CB OD2 CG
BOND N HN N CA C CA C +N
BOND CA HA CB HB1 CB HB2
DOUBLE O C CG OD1
IMPR N -C CA HN C CA +N O
IMPR CG CB OD2 OD1
CMAP -C N CA C N CA C +N
You'll note that the items being skipped are those with -C or +N because
there is no previous or later patch. The CMAP you want should be added by
the ACED patch, but it requires that the CT3 patch also be applied to
create the NT atom. The logic in psfgen applies ACED, then CT3.
Actually, psfgen doesn't appear to do anything with cross-terms in patches
because it's missing those lines, which would cause ACE, ACED, ACP, ACPD,
and CT3 termini to be missing cross-terms. This was an easy fix, and I
have also reversed the order to apply last (CT3) before first (ACED).
The fix will be in the Feb 18 NAMD nightly build standalone psfgen.
Please let me know if that does or does not resolve your problem.
Thank you very much for catching and reporting this!
Jim
On Tue, 17 Feb 2015, Brian Radak wrote:
> Just a quick follow up, apparently this CMAP omission is intentional, as
> evidenced by the error(?) messages:
>
> psfgen) Info: skipping bond C-N at beginning of segment.
> psfgen) Info: skipping improper C-CA-N-O at beginning of segment.
> psfgen) Info: skipping improper N-C-CA-HN at beginning of segment.
> *psfgen) Info: skipping cross-term C-N-CA-C-N-CA-C-N at beginning of
> segment.*
> psfgen) Info: skipping conformation CA-C-N-CA at beginning of segment.
> psfgen) Info: skipping conformation N-CA-C-O at beginning of segment.
> psfgen) Info: skipping conformation N-CA-C-N at beginning of segment.
> psfgen) Info: skipping conformation C-N-CA-C at beginning of segment.
> psfgen) Info: skipping conformation C-CA-N-HN at beginning of segment.
>
> (bold added by me). All of the other messages are apparently consistent
> with CHARMM behavior. Again, I guess this is a confusing case because we
> only have ONE residue and it is also the first/last residue. Any
> suggestions?
>
> Brian
>
>
>
> On Tue, Feb 17, 2015 at 9:42 AM, Brian Radak <brian.radak.accts_at_gmail.com>
> wrote:
>
>> I am making various dipeptides with psfgen, but I also compared the
>> analogous procedure with CHARMM.
>>
>> However, when I check the gas phase energy of the compound, I get a
>> discrepancy in the dihedral energy corresponding to a missing CMAP term in
>> the PSF generated by psfgen (this is obvious when you set "mergeCrossterms
>> no").
>>
>> It looks like this is a tricky case, as the C36 files have special patches
>> just for dipeptides (at least a special ACED patch as opposed to the normal
>> ACE patch). There does not appear to be special CT3 patch for dipeptides
>> (should there be?). Is this something that psfgen was designed to do?
>>
>> You can see this all in one namd configure file provided you adjust some
>> paths below:
>>
>> +++++++++++++++
>> load VMD_INSTALL_LIB_DIR/vmd/plugins/LINUXAMD64/tcl/psfgen1.6/libpsfgen.so
>>
>> paratypeCharmm on
>> mergeCrossterms no
>> topology PATH_TO_CHARMM_PARAMETERS/toppar/top_all36_prot.rtf
>> parameters PATH_TO_CHARMM_PARAMETERS/toppar/par_all36_prot.prm
>>
>> segment PROT {
>> residue 1 ASP
>> first ACED
>> last CT3
>> auto angles dihedrals
>> }
>> # need this to seed the IC table
>> coord PROT 1 N {0.0 0.0 0.0}
>> coord PROT 1 CA {1.48 0.0 0.0}
>> coord PROT 1 C {1.99 1.4 0.0}
>> regenerate angles dihedrals
>> guesscoord
>>
>> set basename foo
>> writepsf ${basename}.psf
>> writepdb ${basename}.pdb
>>
>> exclude scaled1-4
>> switching off
>> cutoff 999
>> outputname $basename
>> structure ${basename}.psf
>> coordinates ${basename}.pdb
>> temperature 0
>>
>> run 0
>> +++++++++++++++
>>
>> --
>> Brian Radak
>> Postdoctoral Scholar
>> University of Chicago
>> Department of Biochemistry & Molecular Biology
>> Gordon Center for Integrative Science, W323A
>> 929 E. 57th St.
>> Chicago, IL 60637-1454
>> Tel: 773/834-2812
>> e-mail: radak_at_uchicago.edu
>>
>
>
>
> --
> Brian Radak
> Postdoctoral Scholar
> University of Chicago
> Department of Biochemistry & Molecular Biology
> Gordon Center for Integrative Science, W323A
> 929 E. 57th St.
> Chicago, IL 60637-1454
> Tel: 773/834-2812
> e-mail: radak_at_uchicago.edu
>
This archive was generated by hypermail 2.1.6 : Tue Dec 27 2016 - 23:20:54 CST