Re: [modeller_usage] restraints loosed after loop refinement

[Modeller Caretaker <>]

mario ciappy wrote:
> I have a problem with restraints after loop refeinement, in particular
> the distances between some atoms do not respect the restraint values set
> up. These values are respected after model building (without a slow MD)
> but not after loop refinements.
> I didn't use any refinement of imput model.
> 
> I show the used scripts:

For your automodel script, you create a MyModel class which adds your
custom restraints. But for some reason you do not do the same thing when
you use loopmodel. Thus, Modeller will know nothing about those
restraints and will not honor them.

Something like the following will probably work for your loop
refinement; the "special_restraints" method is used both for comparative
modeling and for refinement:

log.verbose()
env = environ()

class MyLoopModel(loopmodel):
    def special_restraints(self, aln):
        rsr = self.restraints
        at = self.atoms

        rsr.add(forms.gaussian(group=physical.xy_distance,
                               feature=features.distance(at['NZ:114'],
                                                         at['OD1:288']),
                               mean=2.6, stdev=0.1))

        rsr.add(forms.gaussian(group=physical.xy_distance,
                               feature=features.distance(at['NE:136'],
                                                         at['OD1:260']),
                               mean=2.6, stdev=0.3))

a = MyLoopModel(env,
              alnfile  = 'allinea.pir',
              knowns   = ('AB', 'BC'),
              sequence = 'CD')
a.starting_model = a.ending_model = 1
a.md_level = None
a.loop.starting_model = a.loop.ending_model = 1
a.loop.md_level = refine.fast
a.make()


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