AMBER Archive (2009)

Subject: Re: [AMBER] Attaching a ligand to a protein residue.

From: Kipras Redeckas (kiprasr_at_gmail.com)
Date: Sun Sep 27 2009 - 05:23:42 CDT


2009/9/26 case <case_at_biomaps.rutgers.edu>

> On Sat, Sep 26, 2009, Kipras Redeckas wrote:
> >
> > I'm currently working on bacteriorhodopsin and I'm having some trouble
> with
> > it. I can't seem to find a good way of creating Amber topology and
> parameter
> > files of it. Bacteriorhodopsin has its ligand (retinal) covalently bound
> to
> > a lysine residue via a Shiff-base link. I extracted the ligand from the
> pdb
> > file, added hydrogens to it with Gaussian and used antechamber to create
> > prep and frcmod files from the Gaussian output. This way retinal simply
> > appeared as a non-bound residue after I ran xleap. I tried deleting the
> > excess hydrogens on both sodium (from lysine) and carbon (from retinal)
> and
> ^^^^^^^
>
> In English, N is "nitrogen", (sodium is "Na").
>
> > creating a bond between them with xleap but that didn't work. The
> tutorial I
> > found here http://ambermd.org/tutorials/advanced/tutorial1/ seems to
> cope
> > with a similar problem but I had no luck in adapting these methods to
> mine.
>
> Indeed, the tutorial should provide a good pathway. You want to create a
> single "residue" that consists of the lysine that has the Schiff base
> connection to the retinal -- you don't want two residues (LYS and RET), but
> rather one in which the nitrogen-carbon bond is already made. You can then
> process this through antechamber.
>
> Beyond that, it's hard to be of much help without knowing more details of
> what
> you actually did.
>
> ....dac
>
>
> _______________________________________________
> AMBER mailing list
> AMBER_at_ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>

Yeah, I made a mistake about sodium and nitrogen, these two got kinda mixed
up in my head :)

I tried your suggestion and with a bit of improvisation of my own I think
that I may have got my results. This is how it went:
*I created a single protonated residue out of RET and LYS.
*I ran it through antechamber and obtained the missing parameters.
*Now all that was missing were the parameters of the amino acid connections,
because antechamber changed LYS atoms to gaff type (lower case letters).
*So I simply extracted the missing parameters from parm99.dat and included
them into frcmod file with letter case changes.
*Now Amber was able construct the input parameters.

I tried a different approach because I have the Schiff base parameters from
several publications. After antechamber generated its results I used saveoff
in tleap to create a library file. In that file I manually changed the types
of lysine atoms in the conceived residue to their corresponding Amber types
of regular lysine (lower case to upper case). Now all I needed were the
parameters concerning the RET-LYS link (the one that now had a gaff-Amber
connection). I simply wrote these down from the publications. Both of my
attempts seemed to yield the same initial results.

Everything seems to work fine, although I'm concerned whether my approach of
this problem is accurate. Mainly because I've never done anything like that
before. :)
_______________________________________________
AMBER mailing list
AMBER_at_ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber