Hi Juan

I think you need some help from GUI tool (VMD etc) to do so. If so the way is to know the amino acids ~5-6 angstroms surrounding to your ligand in the second PDB file. Thereafter, from the alignment build a chimera model by saying to copy ligand and coordinates of the amino acids ~5-6 angstroms surrounding to your ligand in your second .pdb file (with deleted all atom coordinates except ligand and ~5-6 angstroms residues surrounding to it) and rest from first template.

Is that not possible in your case to build model only based on first template and copy ligand simply through superposition of the homology model on to your second .pdb that contains ligand. After doing so I suggest simple energy minimization in GROMACS or any package to prevent any atom clashes that arised while superposition.

I also await for Benn's reply and analysis on this steps.

Thank you

On Thu, Feb 6, 2014 at 5:49 PM, Juan Munoz-Garcia <juan.munoz-garcia@bioch.ox.ac.uk> wrote:
Dear Mr. Webb,

thank you for your reply. I'm a new MODELLER user so I don't fully understand how should I modify the alignment
as you suggest. Could you give me a example of it?

Thank you.
Juan C. Munoz-Garcia
From: Modeller Caretaker [modeller-care@salilab.org]
Sent: 05 February 2014 17:26
To: Juan Munoz-Garcia; modeller_usage@salilab.org
Subject: Re: [modeller_usage] Using 2 templates, one to just model the protein and the other to just model the ligand

On 02/05/2014 02:03 AM, Juan Munoz-Garcia wrote:
> I want to do homology modelling of a protein based on just one template
> because it is so far the only template of a particular protein state of
> its family I' m interest in. However, that template does not contain the
> natural ligand of the target. Then, I want to use another template of
> the protein-ligand(natural) complex of interest but just to say Modeller
> to use the ligand coordinates. So, it is a multi-alignment using the HET
> option but I don't know how to say Modeller that I just want to model
> the target sequence based on just one template and considering the
> ligand of the other template. Is this possible?

This is simply done by modifying the alignment to align the first
template with the model in the "protein" region, and the second template
with the model in the "ligand" region. Note, however, that you can't use
*just* the ligand from one template - you'll have no constraints on it
and so it will likely wander away from the protein during the
optimization. You should try to incorporate at least some of the nearby
protein from the second protein, so that Modeller can deduce
protein-ligand restraints.

        Ben Webb, Modeller Caretaker
modeller-care@salilab.org             http://www.salilab.org/modeller/
Modeller mail list: https://salilab.org/mailman/listinfo/modeller_usage

modeller_usage mailing list

Lalith Kumar K
Research Associate,
CADD Division,
Dr. Reddy's Institute of Life Sciences (Formerly known as Institute of Life Sciences)
University of Hyderabad Campus
Gachibowli, Hyderabad 500046, India
Tel: +91 9502074040