I agree, the way we are doing atom types in the SB protein models is not
going to be easily transferable to automatic atom type assignment.
On Oct 13, 2016 12:11 PM, "David A Case" <david.case.rutgers.edu> wrote:
> On Thu, Oct 13, 2016, Miha Purg wrote:
> >
> > I was trying to use parmchk2 on a non-standard residue (see attached
> > .prepin) that contains a 'CX' atom type...
> >
> > "parmchk2 -i FGly_deprot.prepin -f prepi -o FGly_deprot.frcmod -a Y -p
> > $AMBERHOME/dat/leap/parm/parm10.dat"
> >
> > ...and I received the following error:
> > "Atom type of CX does not shown up in PARMCHK.DAT"
> >
> > I manually changed the atom type of CA from 'CT' (obtained with
> antechamber
> > -at amber) to 'CX' to be consistent with ff14SB. Has PARMCHK.DAT not been
> > updated for ff14SB or am I missing something?
>
> Yes: the antechamber notion of "amber atom type" has not been updated. The
> specific problem of changes in FF14SB can be addressed, and we will try to
> do
> that. The more general problem is tougher: what constitutes an "Amber atom
> type" is very much a moving and variable target. Both the nucleic acid
> community (e.g. bsc1) and the protein people (e.g. in ff15ipq) are adding
> atoms types right and left.
>
> The parmchk2 program was primarily designed to estimate force constants
> for gaff and gaff2 atom types, which are under antechamber's control. You
> can (try to) use the program with Amber atom types, but you should be
> triply careful to check the results...in most cases, it will be better
> to do this job by hand. Note *especially*: if your modified residue has
> chemical functionality that is different from that in the standard amino
> acids and nucleotides supported by Amber atom types, then Amber atoms
> types are not the appropriate universe to be working in.
>
> [cc-ing this Junmei. I'm wondering if we should really be trying to
> support the "-p" option in parmchk2. Maybe we could expand the options
> for the "-s" flag to a restricted set of Amber parameter files, and not
> try to allow an arbitrary file to be read in via the "-p" option.]
>
> Thanks for the report.
>
> ....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 Thu Oct 13 2016 - 09:30:02 PDT
