Hi,
Thanks for the tips. I've found that
http://ffamber.cnsm.csulb.edu/ffamber.php reports the validations of the
GROMACS ports were done with AMBER 8.0. However LYN charges were patched
after 8.0 at
http://ambermd.org/bugfixes/8.0/bugfix.19, so perhaps either
the validation wasn't done with the patched version, or LYN wasn't in the
validation. The HIP change might have been mentioned in AMBER 9.0 release
notes (if there are any?) but they don't seem to be in a bug fix anywhere.
C'est la vie.
I've proposed both the LYN and HIP fixes to go into the version of amber03
in the GROMACS distribution.
Mark
On Tue, Sep 6, 2016 at 2:45 AM David A Case <david.case.rutgers.edu> wrote:
> On Mon, Sep 05, 2016, Mark Abraham wrote:
> >
> > Thanks for the response. It seems that amber03 had LYN charge parameters
> > that changed at some point in the AMBER program prep files, perhaps as a
> > fix. But it is hard for me to find the reason for the change, to know
> > whether e.g. the GROMACS implementation is wrong because we copied the
> old
> > version and haven't known to stay in sync. This apparently does affect
> LYN
> > in amber03, but maybe other things got fixed in AMBER alongside that fix,
> > and this would also need attention in GROMACS. If there was a publicly
> > readable source code repository then I could search it, but I don't think
> > that exists.
>
> In July of 2004(!), there was a commit to the all_amino03.lib file
> regarding
> the LYN charges. The ones originally there were copied from the ff94 force
> field, and had not been updated to reflect the "ff03" way of obtaining
> charges. The 2004 commit provided the "correct" ff03 charges for LYN.
> There haven't been any further changes to the LYN charges.
>
> [FWIW: the only other change I see is a change to the HIP charges in Nov.
> 2005--logs are unclear about why this change was made. As Carlos
> suggested,
> you might want to try to contact Yong Duan directly.
>
> ...hope this helps....dac
>
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Sep 06 2016 - 07:30:02 PDT
