Is there anything wrong with my configuration file?My models look so strange
Hi all, Two days ago I wanted to model an antibody with specifying the CDR regions.Here is my configuation file:
# Homology modeling by the automodel class from modeller import * # Load standard Modeller classes from modeller.automodel import * # Load the automodel class
# Redefine the special_patches routine to include the additional disulfides # (this routine is empty by default): class mymodel(loopmodel): def special_patches(self, aln): # A disulfide between residues 22 and 96: self.patch(residue_type='DISU', residues=(self.residues['22'], self.residues['95'])) self.patch(residue_type='DISU', residues=(self.residues['158'], self.residues['228'])) def select_loop_atoms(self): return selection(self.residue_range('26','33'),self.residue_range('51','57'),self.residue_range('96','110'),self.residue_range('162','172'),self.residue_range('190','192'),self.residue_range('229','237'))
log.verbose() # request verbose output env = environ() # create a new MODELLER environment to build this model in
# directories for input atom files env.io.atom_files_directory = './:../atom_files'
a = mymodel(env, alnfile = '3.ali', # alignment filename knowns = ('HL','y','121','113'), # codes of the templates sequence = '6b4', loop_assess_methods=assess.DOPE) # code of the target a.starting_model= 1 # index of the first model a.ending_model = 8 # index of the last model # (determines how many models to calculate) a.loop.starting_model=1 a.loop.ending_model=2
a.make() # do the actual homology modeling
I don't know what is wrong with it but the models it built were so strange,I can see some breakpoints in the structures.
However,if I delete the sentences for loops specifying,the breakpoints would disappear.
# Homology modeling by the automodel class from modeller import * # Load standard Modeller classes from modeller.automodel import * # Load the automodel class
# Redefine the special_patches routine to include the additional disulfides # (this routine is empty by default): class mymodel(loopmodel): def special_patches(self, aln): # A disulfide between residues 22 and 96: self.patch(residue_type='DISU', residues=(self.residues['22'], self.residues['95'])) self.patch(residue_type='DISU', residues=(self.residues['158'], self.residues['228'])) log.verbose() # request verbose output env = environ() # create a new MODELLER environment to build this model in
# directories for input atom files env.io.atom_files_directory = './:../atom_files'
a = mymodel(env, alnfile = '3.ali', # alignment filename knowns = ('HL','y','121','113'), # codes of the templates sequence = '6b4') a.starting_model= 1 # index of the first model a.ending_model = 8 # index of the last model # (determines how many models to calculate) a.make() # do the actual homology modeling
So could you please tell me how can I specify the CDR loops correctly?Thank you very much for any help.
Best wishes, Stanley
wayj86 wayj86 wrote: > Two days ago I wanted to model an antibody with specifying the CDR > regions.Here is my configuation file: ... > I don't know what is wrong with it but the models it built were so > strange,I can see some breakpoints in the structures.
Your script is a little optimistic - you are asking Modeller to simultaneously optimize 6 loops, some of which are themselves over the sampling limit (12 residues) yet you are building only two loop models for each model, which is nowhere near enough sampling. (For a 12-residue loop, we typically recommend 500 loop models or more for adequate sampling.) In your case, you would probably want to optimize each loop individually, and then combine the best models for each loop.
Ben Webb, Modeller Caretaker
participants (2)
-
Modeller Caretaker
-
wayj86 wayj86