Re: [modeller_usage] Modeling ligands in the binding site
To: Qinghua Liao <>
Subject: Re: [modeller_usage] Modeling ligands in the binding site
From: Modeller Caretaker <>
Date: Mon, 12 Oct 2009 09:21:51 -0700
Cc:
Qinghua Liao wrote:
I want to build a model with a ligand in the binding sites, but I have
problems. I read the manual on-line, but I don't find any python script
for aligning the sequence to the structure with ligand. I don't know the
differences when compared to structure without ligand. Could somebody
give me some example of these python script, or give me some suggestion
on how to edit the python script which is used to align to structure
without ligand? Thanks very much!
Typically you would do a regular alignment of your structure (without
ligand) against the target sequence, then manually edit the alignment
file to add ligand(s). Since ligands are simply copied directly from the
template to the target, the alignment in this region is simply 1:1, so
it is comparatively easy to add manually. And relying on a conventional
sequence alignment algorithm to generate this alignment is asking for
trouble, since most algorithms use a distance matrix (e.g. BLOSUM) that
only contains distances for the standard amino acids, not for the full
range of possible ligands.