March 2010 - modeller_usage

set topology for autobuilder.
by Knut J Bjuland 19 Mar '10

19 Mar '10

I used thesse commando to set toplogy information when doing an alignment. BUt it did work when I wanted to build it. What kind of command should I invoke to add thesse information in the build phase of modeller. env.libs.topology.read(file='$(LIB)/top_allh.lib') env.libs.parameters.read(file='$(LIB)/par.lib') _________________________________________________________________ Hotmail: Trusted email with powerful SPAM protection. https://signup.live.com/signup.aspx?id=60969

2 1

If I disable the default N and C-term patched, must I manually apply them?
by wayj86 wayj86 19 Mar '10

19 Mar '10

Hi all, I hope to model a protein which is only a part of a big one (i.e. not up to its real N or C-terminal). So I turn off the default N and C patches with "patch_default=false". In this case, must I manually apply neutral patches, like ACE and CT3 to my model? I mean, can Modeller work wothout any N and C-term patches? I asked this because my next step is docking, which did not require any preexisting terminal patches. Thank you. Best wishes, Stanley -- Be the change you want to see in the world.

2 1

log files and processing multiple files
by abdullah ahmed 18 Mar '10

18 Mar '10

Dear Modeller users, I would like to ask for help on two problems: 1) I do not get log files after running optimization. I am certain that it is not a matter of them taking really long to be made because I have been using Modeller for 3 months now and I have not gotten a log file. I do however, get "pdb.ini" "pdb.rsr" and "pdb.D000001" (contains info on maximal shift, energy changes and gradients). 2) I have a series of molecules that I would like to apply the same constraints to, and then optimize. Is there a script I can use to optimize all the files in one go? Thanks in advance for the help, Abdullah _________________________________________________________________ Your E-mail and More On-the-Go. Get Windows Live Hotmail Free. https://signup.live.com/signup.aspx?id=60969

2 1

modeller and virtualenv
by Will Dampier 17 Mar '10

17 Mar '10

I've been trying to modeller to play nice with the virtualenv package management system. Has anyone had any success getting this to work? thanks Will

2 1

Model refinement
by Ehab AlMoubarak 16 Mar '10

16 Mar '10

Dear All, I am modelling a potassium channel protein using a template with fairly low sequence similarity to my target, therefore, I am trying to put several restraints. I am struggling to deal with a long loop supposed to be extracellular: so its location to the rest of the protein is roughly known. As I am refining the model using MD, the loop location completely became wrong, I suppose that the absence of alignment in the refinement process is the reason (which covers only small area of the loop anyway). I don’t know if there is anyway to exclude the loop while refining the rest of the model, or to restraint the loop in an approximate location of the structure (like putting distance restraint) while refining it. I must say that the gap in alignment corresponds to the loop is well over 15 amino acids. Any thoughts are highly appreciated. My script is here: from modeller import * from modeller.automodel import * # Load the automodel class log.verbose() # Override the 'special_restraints' and 'user_after_single_model' methods: class MyModel(automodel): def special_restraints(self, aln): # Constrain the A and B chains to be identical (but only restrain # the C-alpha atoms, to reduce the number of interatomic distances # that need to be calculated): s1 = selection(self.chains['A'], self.residue_range('7:A', '30:A'),self.residue_range('107:A', '130:A')).only_atom_types('CA') #s2 = selection(self.chains['A'], self.residue_range('107:A', '130:A')).only_atom_types('CA') s2 = selection(self.chains['A'], self.residue_range('157:A', '180:A'),self.residue_range('217:A', '240:A')).only_atom_types('CA') #s4 = selection(self.chains['A'], self.residue_range('217:A', '240:A')).only_atom_types('CA') s5 = selection(self.chains['A']).only_atom_types('CA') s6 = selection(self.chains['B']).only_atom_types('CA') s7 = selection(self.chains['A'], self.residue_range('80:A', '135:A')).only_atom_types('CA') s8 = selection(self.chains['A'], self.residue_range('185:A', '250:A')).only_atom_types('CA') self.restraints.add(secondary_structure.alpha(self.residue_range('7:A', '30:A'))) #self.restraints.add(secondary_structure.alpha(self.residue_range('107:A', '130:A'))) self.restraints.add(secondary_structure.alpha(self.residue_range('157:A', '180:A'))) #self.restraints.add(secondary_structure.alpha(self.residue_range('217:A', '240:A'))) #self.restraints.add(secondary_structure.alpha(self.residue_range('73:A', '92:A'))) #self.restraints.symmetry.append(symmetry(s1, s2, 1.0)) self.restraints.symmetry.append(symmetry(s7, s8, 1.0)) #self.restraints.symmetry.append(symmetry(s2, s3, 1.0)) #self.restraints.symmetry.append(symmetry(s3, s4, 1.0)) self.restraints.symmetry.append(symmetry(s5, s6, 1.0)) def user_after_single_model(self): # Report on symmetry violations greater than 1A after building # each model: self.restraints.symmetry.report(1.0) env = environ() # directories for input atom files env.schedule_scale = physical.values(default=1.0, soft_sphere=0.7) env.io.atom_files_directory = ['.', '../atom_files'] # Be sure to use 'MyModel' rather than 'automodel' here! a = MyModel(env, alnfile = 'task3Modified8.ali' , # alignment filename knowns = '1orq_9', # codes of the templates sequence = 'TASK3h27', # code of the target assess_methods=assess.DOPE) a.starting_model= 1 # index of the first model a.ending_model = 15 # index of the last model # Very thorough VTFM optimization: a.library_schedule = autosched.slow a.max_var_iterations = 300 # Thorough MD optimization: a.md_level = refine.slow # Repeat the whole cycle 2 times and do not stop unless obj.func. > 1E6 a.repeat_optimization = 2 a.max_molpdf = 1ttt a.make() # Get a list of all successfully built models from a.outputs ok_models = filter(lambda x: x['failure'] is None, a.outputs) # Rank the models by DOPE score key = 'DOPE score' ok_models.sort(lambda a,b: cmp(a[key], b[key])) # Get top model m = ok_models[0] print "Top model: %s (DOPE score %.3f)" % (m['name'], m[key]) My best regards Ehab, Research student

1 0

using the parallel job support
by Will Dampier 16 Mar '10

16 Mar '10

I'm trying to use the parallel job support built into modeller. I've gotten the jobs to work with "local_slave" workers but when I try to use ssh_slave workers I get an OSError "OSError [Errno 2] No such file or directory: '/opt/apps/Protein-Threading' from <ssh_clave on iblade003". I know that ssh_slaves require a shared memory space and I know that everything in properly mounted ... running: ssh iblade003 'ls /opt/apps/Protein-Threading' return the proper results. Any suggestions? Will

2 1

automodel() known templates
by Rafik Ouerchefani 16 Mar '10

16 Mar '10

Hello, I'm following tutorials and the all-steps example. I would like to know if there is a function or a method to automate the selection of the templates to be used in automodel() as "knowns". In the examples, this selection is done manually from matrix.mat or compare.log files. Thanks - Rafik

2 1

How should I fill in the missing rsidues with the same chain id, resnumber and HETATM keeping?
by wayj86 wayj86 16 Mar '10

16 Mar '10

Dear all, I am trying to fix a protein with 6 missing aa. This protein has a resid from 500 to 650, with three HETATM residues (N-glycan) from 800 to 802 ( discontinuous with the protein resnumber). Here is my script to run modeller: *First step: to build a model with blank chain ID and resnumber from 1 to 150:* # Homology modeling by the automodel class from modeller import * # Load standard Modeller classes from modeller.automodel import * # Load the automodel class 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'] # Read in HETATM records from template PDBs env.io.hetatm = True env.patch_default = False class MyModel(automodel): def select_atoms(self): # Select residues 41 and 46 (PDB numbering) return selection(self.residue_range('41', '46')) a = MyModel(env, alnfile = 'xy.pir', # alignment filename knowns = 'x', # sequence = 'y', assess_methods=(assess.DOPE, assess.GA341)) # code of the target a.starting_model= 1 # index of the first model a.ending_model = 10 # index of the last model # Very thorough VTFM optimization: a.library_schedule = autosched.slow a.max_var_iterations = 300 # Thorough MD optimization: a.md_level = refine.slow # Repeat the whole cycle 2 times and do not stop unless obj.func. > 1E6 a.repeat_optimization = 3 a.max_molpdf = 1e6 a.make() # Get a list of all successfully built models from a.outputs ok_models = filter(lambda x: x['failure'] is None, a.outputs) # Rank the models by DOPE score key = 'DOPE score' ok_models.sort(lambda a,b: cmp(a[key], b[key])) # Get top model m = ok_models[0] print "Top model: %s (DOPE score %.3f)" % (m['name'], m[key]) By running this script I did got a model with missing residues fixed and HETATM keeping, but the chain ID was blank (because only one chain was built), and the resnumber was from 1 to 154. I did hope to keep the original chain ID and the resnum, so I try the rename_segments, like: *Second step: reassign the chain ID and resnumber:* # Example for: model.rename_segments() # This will assign new PDB single-character chain id's to all the chains # in the input PDB file (here there are two 'chains': protein and the HETATM # water molecules). from modeller import * # Read the MODEL with all HETATM and water records (so there are two 'chains'): env = environ() env.io.atom_files_directory = ['./', ':../atom_files'] env.io.hetatm = True mdl = model(env, file='y') # Assign new segment names and write out the new model: mdl.rename_segments(segment_ids='A', renumber_residues=[500, 650]) mdl.write(file='y.ini') Now in the final model y.ini the chain ID was defined as "A", and resnum was from 500 to 650. However, the HETATM residues was from 651 to 653, but not 800 to 802. So should I modify my scripts to fix the protein and keeping its original chain ID, resnumber and HETATM residues? Thank you very much for your attention. P.S. The HETATM residues are also in chain A, the same chain with the protein. -- Be the change you want to see in the world.

2 1

Refine segments of a crystal structure
by Baptiste Legrand 16 Mar '10

16 Mar '10

Hi all, Is it possible to refine protein segments from a crystal structure using modeller as we can do for models (but in this case, without providing alignment and template files) ? Thanks. Best wishes, Baptiste

2 1

alignment with modeller
by Knut J Bjuland 16 Mar '10

16 Mar '10

I used multiple alignment with protein sequence to get a better alignment with a template. Is this a good practice. What are the recommende settings in salign to get the best possible alignment.

2 1