AMBER Archive (2009)

Subject: [AMBER] RE: Tutorial 5

From: Ross Walker (ross_at_rosswalker.co.uk)
Date: Fri Jan 16 2009 - 11:10:51 CST

Hi George,

I am copying this message to the amber mailing list since it is the sort of
thing that would interest other amber users as well - see
http://lists.ambermd.org/ for signup details.

> With reference to Tutorial 5
> (http://www.rosswalker.co.uk/tutorials/amber_workshop/Tutorial_five/create
> _prmtop.htm), I have the following query.
>
> If one follows the procedure you describe for the preparation of the
> ligand, the ligand.prepin file contains the internal coordinates. If one
> were to run an md simulation with a given protein in order to calculate
> binding energies and H-bond formation, one would be dealing with two
> different coordinate systems, namely the ligand internal coordinates and
> the protein pdb coordinates. I can't see how to run this type of
> simulation, unless the starting ligand.pdb coordinate file was originally
> taken from a protein/ligand pdb coordinate file.
>
> Am I right in this assumption?

Technically yes but that assumes you don't have some kind of pdb file with
the coordinates in in the first place. The internal coordinates in the prep
or mol2 file are only used if atoms are missing from the actual pdb that is
loaded into leap, otherwise the coordinates in the pdb file are used. So
indeed as you suggest you do need input structures containing compatible
coordinates.

Running the ligand simulation and the protein simulation separately with
different coordinate sets doesn't matter, it is the simulation of the
complex where it presents a problem. In an ideal world you would just place
these two systems in solution near each other and run for an infinite time
and they would explore all the phase space including all of the places where
the ligand can bind and you would get perfectly weighted statistics. Of
course this is not possible and thus you need to provide a structure which
has the ligand coordinated to the protein in some way. The easiest method is
generally to find a crystal structure that contains the ligand. However, if
this is not possible then you can find something similar and manually place
the ligand in the correct binding pocket - any missing atoms get added by
leap based on the internal coordinates in the mol2/prep file.

Alternatively if you don't have information about where the ligand binds
then you would need to run some kind of docking simulation to generate
potential structures for the protein-ligand complex - for example something
like Dock. The pdb from this you would then use with leap (having loaded the
relevant force field files and the mol2/prep file to describe the topology
of your ligand) to produce the prmtop and inpcrd files required by sander.

I hope that makes sense.

All the best
Ross

/\
\/
|\oss Walker

| Assistant Research Professor |
| San Diego Supercomputer Center |
| Tel: +1 858 822 0854 | EMail:- ross_at_rosswalker.co.uk |
| http://www.rosswalker.co.uk | PGP Key available on request |

Note: Electronic Mail is not secure, has no guarantee of delivery, may not
be read every day, and should not be used for urgent or sensitive issues.

_______________________________________________
AMBER mailing list
AMBER_at_ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber