Re: [modeller_usage] A single type of rotamer

[Modeller Caretaker <>]

On 10/8/10 9:38 PM, Omar Piña wrote:
> I need to model a protein with the lysine all straight (stuck-up?). How
> can I say to modeller that only create models with the lysine in that
> form?? This Lysine is known to be like that. Thanks!!

For any modeling in Modeller, you need to think about the set of spatial 
restraints that correspond to your desired geometry. By "stuck-up" I 
presume you mean a specific rotamer for which you know chi1, chi2 etc. 
So you just create a few restraints on those dihedrals. If you're doing 
this during a comparative modeling run with automodel, you'd build those 
restraints in the special_restraints() routine. The only complication is 
that Modeller already builds homology-derived sidechain restraints, so 
you'd have to remove those first using Restraints.unpick() and 
Restraints.condense().

I've attached a simple modification of the basic automodel example from 
http://salilab.org/modeller/9v8/manual/node16.html which modifies it to 
restrain the first Lys residue (residue 15) to have both chi1 and chi2 
be 0 +/ 0.1 radians.

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

from modeller import *
from modeller.automodel import *

log.verbose()
env = environ()

env.io.atom_files_directory = ['.', '../atom_files']

class MyModel(automodel):
    def special_restraints(self, aln):
        rsr = self.restraints
        lys = self.residues['15']
        for d in lys.chi1, lys.chi2:
            rsr.unpick(*d.atoms)
        rsr.condense()
        rsr.add(forms.gaussian(group=physical.chi1_dihedral,
                               feature=features.dihedral(*lys.chi1.atoms),
                               mean=0.0, stdev=0.1))
        rsr.add(forms.gaussian(group=physical.chi2_dihedral,
                               feature=features.dihedral(*lys.chi2.atoms),
                               mean=0.0, stdev=0.1))

a = MyModel(env, alnfile='alignment.ali', knowns='5fd1', sequence='1fdx')
a.starting_model = a.ending_model = 1
a.make()