Dear users, I would like to build part of a protein which is missing from the PDB file. It is a very long loop (72 residues) and I am not sure if the task is beyond Modeller skills. I have information about secondary structure elements of this protein part (three helices and a strand) thus I tried to use them as restrains in the program but what I got after increasing the automodel.max_molpdf (because it did not produce any results with the default limit) was just a long flat and linear loop without any apparent secondary structure. Might be I did not check enough into the available literature about the program and there are ways to model such a loop or it is just not possible. Can someone, please, help me with a suggestion?

that's the input file:

# Homology modelling by the automodel class with special restrains for secondary structure

from modeller.automodel import * # Load the automodel class

# Redefine the special_restraints routine to include the secondary # structure restraints (this routine is empty by default): class mymodel(automodel): def special_restraints(self, aln): rsr = self.restraints # Three alpha-helix: rsr.make(aln, restraint_type='ALPHA', residue_ids=('143', '148'), spline_on_site=False) rsr.make(aln, restraint_type='ALPHA', residue_ids=('195', '200'), spline_on_site=False) rsr.make(aln, restraint_type='ALPHA', residue_ids=('204', '206'), spline_on_site=False) # One strand: rsr.make(aln, restraint_type='STRAND', residue_ids=('171', '174'), spline_on_site=False)

log.verbose() env = environ()

# directories for input atom files env.io.atom_files_directory = './'

a = mymodel(env, alnfile='protein_full.ali', # alignment filename knowns='PDB_refined', # codes of the templates sequence='protein_full') # code of the target

a.starting_model= 1 # index of the first model a.ending_model = 3 # index of the last model # (determines how many models to calculate)

a.max_molpdf = 500000 # define the max of molec probability density function

a.md_level = refine.fast # type of refinement of model

a.make()

thanks loris...