[gmx-users] histidine (HISB) planarity

[Berk Hess]

>From: Anthony Armstrong <armstron at queen.med.jhmi.edu>
>Reply-To: Discussion list for GROMACS users <gmx-users at gromacs.org>
>To: gmx-users at gromacs.org
>Subject: [gmx-users] histidine (HISB) planarity
>Date: Fri, 11 Aug 2006 17:19:24 -0400 (EDT)

>
>I was having problems keeping my histidines planar, and had to manually
>edit the .top file produced by pdb2gmx in order to fix the problem.  I was
>just wondering if this problem (i.e. incorrect and incomplete improper
>dihedral records in the top file) is isolated to histidine topologies
>generated by pdb2gmx using the G53a6 ff parameters or if manual
>intervention is frequently required for such standard geometries.  I'm
>using Gromacs v3.3.  I start with a pdb file of an extended
>peptide (with planar his) and issue:
>
>pdb2gmx -f -ignh -inter 1STN.pdb -o 1STN.gro -p 1STN.top -ff G53a6
>
>whereupon I specify the single histidine to be of type HISB. After:

>...puckered...
>
>To correct this it was necessary to delete:
>    32    35    30    33     2    gi_1
>(this auto generated improper was largely responsible for the ring
>puckering - its removal alone caused much of the problem to be corrected)
>
>and to add:
>    34    30    35    33     2    gi_1
>
>(which brough the HE2 hydrogen slightly out of plain back into plain)
>

Indeed this is incorrect.
Atom number 32 in the improper dihedral should be replaced by 34.
The error is caused by a mistake in ffG53a5/6.rtp fro HISB, the line:
CE1   ND1   NE2   HD2     gi_1
should be:
CE1   ND1   NE2   HE1     gi_1

I have corrected this for the next release.

>Is it recommended that I look over every record in the topology file?  Is
>this necessary?  Thanks.

Well, it is always recommended to check to whole topology, as one can
never exclude errors. However, for a reasonable sized protein this
is a nearly impossible task.
The hope is that when the force field files, in this case the building block
files, do not change too often and too much and when thousands of users
use them (like with Gromacs), errors will be found quickly and files should
be reliable.
In this case the error stems for the original GROMOS building block file
we have used for making the Gromacs rtp files. The previous GROMOS
43 and 45 force fields did not have explicit hydrogens on the carbons
in the ring and apparently the improper for one of the new hydrogens
was typed incorrectly.
All other interactions for all histidine types seem to be correct.

Thanks for reporting this error,

Berk.