- modeller_usage - salilab.org

loop modeling and hydrogen atoms
by flavio seixas 25 Jul '16

25 Jul '16

Hi, I generated a model with all hydrogens using allhmodel. Now I want to refine a loop in this model using the script below. However, the hydrogens are being removed in the generated models. Would you help me? from modeller import * from modeller.automodel import * log.verbose() env = environ() env.io.atom_files_directory = ['.', '../atom_files'] # Read in HETATM records from template PDBs env.io.hetatm = True env.io.water = True env.io.hydrogen = True # Create a new class based on 'loopmodel' so that we can redefine # select_loop_atoms class MyLoop(loopmodel): # This routine picks the residues to be refined by loop modeling def select_loop_atoms(self): # Two residue ranges (both will be refined simultaneously) return selection(self.residue_range('399:', '406:')) m = MyLoop(env, inimodel='Model172.pdb', sequence='protein', loop_assess_methods=(assess.DOPE, assess.GA341)) m.loop.starting_model= 1 # index of the first loop model m.loop.ending_model = 200 # index of the last loop model m.loop.md_level = refine.very_fast # loop refinement method m.make() # Get a list of all successfully built models from m.loop.outputs ok_models = filter(lambda x: x['failure'] is None, m.loop.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 1: %s (DOPE score %.3f)" % (m['name'], m[key]) m = ok_models[1] print "Top model 2: %s (DOPE score %.3f)" % (m['name'], m[key]) m = ok_models[2] print "Top model 3: %s (DOPE score %.3f)" % (m['name'], m[key]) m = ok_models[3] print "Top model 4: %s (DOPE score %.3f)" % (m['name'], m[key]) m = ok_models[4] print "Top model 5: %s (DOPE score %.3f)" % (m['name'], m[key]) ------------------------------------- Flavio Augusto Vicente Seixas Laboratory of Structural Biochemistry Department of Biochemistry Universidade Estadual de Maringá, PR, Brazil http://www.uem.br

2 1

Why there is a DOPE score decreases when I optimize the selected atoms of automodel?
by Mahesh Velusamy 15 Jul '16

15 Jul '16

Dear All, I have tried generate a models for single point mutation along with optimization and refinement on the selected region of atoms but I am getting bad DOPE score(lesser negative value -11k) for optimized models compared to unoptimized automodels DOPE score(-36K). Also, I observed that DOPE score get increase when I increase the number of atoms for optimization but at same time this leads to bad models. May I know what is the reason for this issue? because when I do the same thing with mutate_model.py I got the same DOPE score as automodels. Here, I am giving the MyModel class where I define my own optimization,refinement and restraints criteria (same as mutate_model.py): class MyModel(automodel): def select_atoms(self): # a non-bonded pair has to have at least as many selected atoms self.env.edat.nonbonded_sel_atoms = 1 sched = autosched.loop.make_for_model(self) s=selection( self.atoms['CA:'+'102'+':'+'A'].select_sphere(20)).by_residue() self.restraints.unpick_all() self.restraints.pick(s) s.energy() #s.randomize_xyz(deviation=-4) s.energy() return s def optimize(s, sched): # conjugate gradient for step in sched: step.optimize(s, max_iterations=200, min_atom_shift=0.001) # md refine(s) cg = conjugate_gradients() cg.optimize(s, max_iterations=200, min_atom_shift=0.001) def refine(self,atmsel,actions): # at T=1000, max_atom_shift for 4fs is cca 0.15 A. md = molecular_dynamics(cap_atom_shift=0.39, md_time_step=4.0, md_return='FINAL') init_vel = True for (its, equil, temps) in ((200, 20, (150.0, 250.0, 400.0, 700.0, 1000.0)), (200, 600, (1000.0, 800.0, 600.0, 500.0, 400.0, 300.0))): for temp in temps: md.optimize(atmsel, init_velocities=init_vel, temperature=temp, max_iterations=its, equilibrate=equil) init_vel = False def special_restraints(self, aln): rsr = self.restraints rsr.clear() s = selection(self) for typ in ('stereo', 'phi-psi_binormal'): rsr.make(s, restraint_type=typ, aln=aln, spline_on_site=True) for typ in ('omega', 'chi1', 'chi2', 'chi3', 'chi4'): rsr.make(s, restraint_type=typ + '_dihedral', spline_range=4.0, spline_dx=0.3, spline_min_points=5, aln=aln, spline_on_site=True) Thanking you in advance With Regards Mahesh V

2 1

parallel modeling _automodel class
by Sajad Ahrari 15 Jul '16

15 Jul '16

Dear users, I am trying to do a parallel job for protein loop refinement by MODELLER9.17. I prepared the scripts as follows and provided all the input files in my working directory and I am running MODELLER through python2.7 (by typing: "python ModellerModelling.py" ). Although I encounter no error through the modeling process, but I doubt if I am doing a parallel job since when I check my workstation CPU usage, only one CPU is engaged (the usage percent does not exceed 100%). Could you please help me know what am I missing? Best regards,Sajjad ModellerModeling.py: # --- UCSF Chimera Copyright --- # Copyright (c) 2000 Regents of the University of California. # All rights reserved. This software provided pursuant to a # license agreement containing restrictions on its disclosure, # duplication and use. This notice must be embedded in or # attached to all copies, including partial copies, of the # software or any revisions or derivations thereof. # --- UCSF Chimera Copyright --- # This script is generated by the Modeller Dialog in Chimera, # incorporating the settings from the user interface. User can revise # this script and submit the modified one into the Advanced Option panel. # Import the Modeller module from modeller import * from modeller.automodel import * from modeller.parallel import * from mymodel import MyModel log.minimal() a.use_parallel_job(j) # Use the job for model building # Use 16 CPUs in a parallel job on this machine j = job() j.append(local_slave()) j.append(local_slave()) j.append(local_slave()) j.append(local_slave()) j.append(local_slave()) j.append(local_slave()) j.append(local_slave()) j.append(local_slave()) j.append(local_slave()) j.append(local_slave()) j.append(local_slave()) j.append(local_slave()) j.append(local_slave()) j.append(local_slave()) j.append(local_slave()) j.append(local_slave()) results = j.run_all_tasks() print("Got model resolution: " + str(results)) mymodel.py: from modeller import * from modeller.automodel import * # ---------------------- namelist.dat -------------------------------- # A "namelist.dat" file contains the file names, which was output from # the Modeller dialog in Chimera based on user's selection. # The first line it the name of the target sequence, the remaining # lines are name of the template structures namelist = open( 'namelist.dat', 'r' ).read().split('\n') tarSeq = namelist[0] template = tuple( [ x.strip() for x in namelist[1:] if x != '' ] ) # ---------------------- namelist.dat -------------------------------- # This instructs Modeller to display all log output. log.verbose() # create a new Modeller environment env = environ() # Directory of atom/PDB/structure files. It is a relative path, inside # a temp folder generated by Chimera. User can modify it and add their # absolute path containing the structure files. env.io.atom_files_directory = ['.', './template_struc'] # Read in HETATM records from template PDBs env.io.hetatm = True # Read in water molecules from template PDBs env.io.water = True # create a subclass of automodel or loopmodel, MyModel. # user can further modify the MyModel class, to override certain routine. class MyModel(loopmodel): def select_loop_atoms(self): from modeller import selection return selection( self.residue_range('256', '309')) def select_atoms(self): from modeller import selection return selection( self.residue_range('256', '309')) def customised_function(self): pass #code overrides the special_restraints method #def special_restraints(self, aln): #code overrides the special_patches method. # e.g. to include the addtional disulfides. #def special_patches(self, aln): a = MyModel(env, sequence = tarSeq, # alignment file with template codes and target sequence alnfile = 'alignment.ali', # name of initial PDB template knowns = template[0]) # one fixed model to base loops on a.starting_model = 1 a.ending_model = 1 # 1 loop models loopRefinement = True a.loop.starting_model = 1 a.loop.ending_model = 1000 a.loop.assess_methods=(assess.DOPE, assess.GA341, assess.normalized_dope) # Assesses the quality of models using # the DOPE (Discrete Optimized Protein Energy) method (Shen & Sali 2006) # and the GA341 method (Melo et al 2002, John & Sali 2003) a.assess_methods = (assess.GA341, assess.normalized_dope) # ------------------------- build all models -------------------------- a.make() # ---------- Accesing output data after modeling is complete ---------- # Get a list of all successfully built models from a.outputs if loopRefinement: ok_models = filter(lambda x: x['failure'] is None, a.loop.outputs) else: ok_models = filter(lambda x: x['failure'] is None, a.outputs) # Rank the models by index number #key = 'num' #ok_models.sort(lambda a,b: cmp(a[key], b[key])) def numSort(a, b, key="num"): return cmp(a[key], b[key]) ok_models.sort(numSort) # Output the list of ok_models to file ok_models.dat fMoutput = open('ok_models.dat', 'w') fMoutput.write('File name of aligned model\t GA341\t zDOPE \n') for m in ok_models: results = '%s\t' % m['name'] results += '%.5f\t' % m['GA341 score'][0] results += '%.5f\n' % m['Normalized DOPE score'] fMoutput.write( results ) fMoutput.close()

2 1

What is a difference between automodel and loopmodel with respect to refining only part of the model optimization ?
by Mahesh Velusamy 07 Jul '16

07 Jul '16

Dear All, Let me explain that what I have tried: I have generated 100 models for one single point mutation sequence following optimization(loop model) on single point mutated residue along with neighbour residues(with different sphere range 1A°-10A°). Now, the observation from my out put is as following: 1. When i optimized the SPM model with the large sphere (10A°) of neighbour residues, I got bad model compared to crysal structure but DOPE score was really high(~5K). 2.When i optimized the SPM model with the less sphere (1A°) of neighbour residues, I got similar model compared to crysal structure but DOPE score was really bad(~250). So , I dont understand why I am getting less DOPE score when I have similar good models compared to crystal structure and vice versa. Then, I followed the same procedure but automodel default optimization on selceted region where i got all the models simlar to crystal structure with less difference between dope score(range ~5k to 6K) Now, I want to what is the difference between automodel and loopmodel optimization on selected region ? and what loop model optimization exactly does ? and why there is a illogical contrast between DOPE score in loop model optimization? Thanking you in advance With Regards Mahesh V

2 1

Modeller 9.17 release
by Modeller Caretaker 06 Jul '16

06 Jul '16

The new version of Modeller, 9.17, is now available for download! Please see the download page at https://salilab.org/modeller/ for more information. If you have a license key for Modeller 8 or 9, there is no need to reregister for Modeller 9.17 - the same license key will work. (It won't do any harm to reregister if you want to, though!) 9.17 is primarily a bugfix release relative to the last public release (9.16). Major user-visible changes include: # loopmodel-generated PDB or mmCIF files now contain extra header information, including the values of any assessment scores. # automodel-generated PDB or mmCIF files now list, for each template, the sequence identity and the range of residues modeled. # EM density maps used for purposes other than density.grid_search() (e.g. use in Flex-EM) no longer have to be cubic. # Chain breaks are now shown (if appropriate) when showing mismatches between a sequence and a structure (SequenceMismatchError). See the Modeller manual for a full change log: https://salilab.org/modeller/9.17/manual/node40.html If you encounter bugs in Modeller 9.16, please see https://salilab.org/modeller/9.17/manual/node10.html for information on how to report them. Note: you are receiving this email either because you subscribed to the modeller_usage mailing list, or you provided this address when you requested a Modeller license and you ticked the "notify me of new releases" box. In the latter case, if you no longer wish to receive announcements of new Modeller releases, simply reply to this email and let us know. Ben Webb, Modeller Caretaker -- modeller-care(a)salilab.org https://salilab.org/modeller/ Modeller mail list: https://salilab.org/mailman/listinfo/modeller_usage

1 0

is it possible run MUTATE_MODEL (optimization step alone) in multithreading / Parallel job ?
by Mahesh Velusamy 05 Jul '16

05 Jul '16

Dear All, Actually, I am using Mutate_model.py script(http://salilab.org/modeller/wiki/Mutate%20model) to generate single point mutant models for large mutation library with an optimization of selected region(~280 atoms). There is no issue on time using for mutating residues but the script takes long time to finish the optimization step alone. So, I want to put this optimization step alone as a multithreading (running on all the CPU cores)similar to a.use_parallel_job(j)# Use the job for model building If it is possible kindly some one could help me out of this would be really useful Thanking you in advance With Regards Mahesh V

2 1

Doubts regrading implementing parallel job on mutate_model.py optimization step
by Mahesh Velusamy 05 Jul '16

05 Jul '16

Dear All, I have tried to implement parallel job on mutate model optimization steps where I selected 457 atoms for optimization (I am doing this because of implement our own objective) and to do this I added the following lines in Mutate_Model.py: from modeller.parallel import * j = job() j.append(local_slave()) j.append(local_slave()) j.append(local_slave()) j.append(local_slave()) j.queue_task(optimize(s,sched)) j.queue_task(refine(atmsel)) j.run_all_tasks() I am not sure I am doing this in correct way but I am getting the following errors: Running 4 tasks on 4 slaves /usr/bin/python2.7 /usr/lib/modeller9.14/bin/modslave.py -slave 172.16.35.74:48622:0:TPLDZNFC >& mutate_model.slave0 /usr/bin/python2.7 /usr/lib/modeller9.14/bin/modslave.py -slave 172.16.35.74:48622:1:INNJBKXG >& mutate_model.slave1 /usr/bin/python2.7 /usr/lib/modeller9.14/bin/modslave.py -slave 172.16.35.74:48622:2:XZYKJCXC >& mutate_model.slave2 /usr/bin/python2.7 /usr/lib/modeller9.14/bin/modslave.py -slave 172.16.35.74:48622:3:XJXTJMBR >& mutate_model.slave3 Traceback (most recent call last): File "/comptes/invite-u/PycharmProjects/untitled/3O26_050716_144858/R12H_A102C_F104A/mutate_model.py", line 104, in <module> j.run_all_tasks() File "/usr/lib/python2.7/dist-packages/modeller/parallel/job.py", line 156, in run_all_tasks for task in self._finish_all_tasks(): File "/usr/lib/python2.7/dist-packages/modeller/parallel/job.py", line 189, in _finish_all_tasks task = self._process_event(obj, s) File "/usr/lib/python2.7/dist-packages/modeller/parallel/job.py", line 201, in _process_event slave.run_task(self.tasks.pop(0)) File "/usr/lib/python2.7/dist-packages/modeller/parallel/slave.py", line 52, in run_task for transfer in task.input_files: AttributeError: 'NoneType' object has no attribute 'input_files' Identified slave <Slave on localhost> My SCRIPT: ---------- import sys import os from modeller import * from modeller.optimizers import molecular_dynamics, conjugate_gradients from modeller.automodel import autosched from modeller.parallel import * # Use 4 CPUs in a parallel job on this machine j = job() j.append(local_slave()) j.append(local_slave()) j.append(local_slave()) j.append(local_slave()) def optimize(atmsel, sched): for step in sched: step.optimize(atmsel, max_iterations=200, min_atom_shift=0.001) ########################################### j.queue_task(refine(atmsel)) ############################################## #refine.use_parallel_job(j) cg = conjugate_gradients() cg.optimize(atmsel, max_iterations=200, min_atom_shift=0.001) def refine(atmsel): md = molecular_dynamics(cap_atom_shift=0.39, md_time_step=4.0, md_return='FINAL') init_vel = True for (its, equil, temps) in ((200, 20, (150.0, 250.0, 400.0, 700.0, 1000.0)), (200, 600, (1000.0, 800.0, 600.0, 500.0, 400.0, 300.0))): for temp in temps: md.optimize(atmsel, init_velocities=init_vel, temperature=temp, max_iterations=its, equilibrate=equil) init_vel = False def make_restraints(mdl1, aln): rsr = mdl1.restraints rsr.clear() s = selection(mdl1) for typ in ('stereo', 'phi-psi_binormal'): rsr.make(s, restraint_type=typ, aln=aln, spline_on_site=True) for typ in ('omega', 'chi1', 'chi2', 'chi3', 'chi4'): rsr.make(s, restraint_type=typ+'_dihedral', spline_range=4.0, spline_dx=0.3, spline_min_points = 5, aln=aln, spline_on_site=True) log.verbose() env = environ(rand_seed=int(-4000)) env.io.hetatm = True env.edat.dynamic_sphere=False env.edat.dynamic_lennard=True env.edat.contact_shell = 4.0 env.edat.update_dynamic = 0.39 env.libs.topology.read(file='$(LIB)/top_heav.lib') env.libs.parameters.read(file='$(LIB)/par.lib') mdl1 = model(env, file = "3O26") ali = alignment(env) ali.append_model(mdl1, atom_files= "3O26.pdb", align_codes= "3O26") s1=selection(mdl1.chains['A'].residues['104']) s1.mutate(residue_type='ALA') s2=selection(mdl1.chains['A'].residues['102']) s2.mutate(residue_type='CYS') s3=selection(mdl1.chains['A'].residues['12']) s3.mutate(residue_type='HIS') ali.append_model(mdl1, align_codes="3O26") mdl1.clear_topology() mdl1.generate_topology(ali[-1]) mdl1.transfer_xyz(ali) mdl1.build(initialize_xyz=False, build_method='INTERNAL_COORDINATES') mdl2 = model(env, file="3O26.pdb") mdl1.res_num_from(mdl2,ali) mdl1.write(file="3O26"+'ALA'+'104'+'CYS'+'102'+'HIS'+'12'+"A"+'.tmp') mdl1.read(file="3O26"+'ALA'+'104'+'CYS'+'102'+'HIS'+'12'+"A"+'.tmp') make_restraints(mdl1, ali) mdl1.env.edat.nonbonded_sel_atoms=1 sched = autosched.loop.make_for_model(mdl1) s = selection(mdl1.atoms['CA:'+'104'+':'+'A'].select_sphere(10),mdl1.atoms['CA:'+'102'+':'+'A'].select_sphere(10),mdl1.atoms['CA:'+'12'+':'+'A'].select_sphere(10)).by_residue() mdl1.restraints.unpick_all() mdl1.restraints.pick(s) s.energy() s.randomize_xyz(deviation=100) mdl1.env.edat.nonbonded_sel_atoms=2 ####################################### j.queue_task(optimize(s,sched)) ####################################### mdl1.env.edat.nonbonded_sel_atoms=1 ######################################### j.queue_task(optimize(s,sched)) ######################################### s.energy() ####################################### j.run_all_tasks() ####################################"" atmsel = selection(mdl1.chains["A"]) score = atmsel.assess_dope() mdl1.write(file="/comptes/invite-u/PycharmProjects/untitled/3O26_050716_144858/R12H_A102C_F104A/R12H_A102C_F104A.MM9999.pdb") os.remove("3O26"+'ALA'+'104'+'CYS'+'102'+'HIS'+'12'+"A"+'.tmp') Kindly correct me If I am wrong and let me know that how to do this in correct way. Thanking you in advance

1 0

Modelling trans-membrane (TM) regions
by Sushil Mishra 29 Jun '16

29 Jun '16

Dear all: I am quite new to MODELLER and I would like to know if it is possible to to models a protein having with few trans-membrane (TM) regions. Are there such an example or tutorial available? I tried to search archive but, unfortunately, there was nothing that can help me. Is it possible to assign a few TM regions (in the form of restrain or something like that) which would be modeled in a plane. Any suggestions/inputs related to modelling enzyme having TM regions will be appreciated. My system has very poor similarity (< 20 %) in PDB, and also no similarity between TM regions of query and TM regions of proteins in PDB. Best, SK

2 2

Is it possible to rum model_single.py in multiple nodes ?
by Mahesh Velusamy 27 Jun '16

27 Jun '16

Dear All, I know there is a option to run modeller in multiple cores but How can run model_single.py in multiple nodes (My wish is to rum 44 nodes where each node has 16 PROCESSOR). If it is possible could you provide me a examples / scripts. Thanking you in advance With Reagrds Mahesh VELUSMAY

2 1

Is it possible to do single residue optimization for automodels ?
by Mahesh Velusamy 27 Jun '16

27 Jun '16

Dear All, Since, My main objective is to get a better conformational sample for a particular single point mutation, I have generated a 100 number of SPM models through automodel but Since, Automodel uses the distance restraints from the template target structure alignment there is not much variations when comparing with the original structure. So I thought to relax the restrains by doing additional optimization on mutated position alone like we see in mutate_model.py http://salilab.org/modeller/wiki/Mutate%20model. Then, I try to modified the automodel script as same like mutate_model.py but I endup in lot of confusions that 1.what shoud I keep for : #conjugate gradient for step in sched: step.optimize(atmsel, max_iterations=200, min_atom_shift=0.001 2.Whether should I include make restraints function ? 3. Is it Ok If I do only optimize and refine step ? def optimize(atmsel, sched): #conjugate gradient for step in sched: step.optimize(atmsel, max_iterations=200, min_atom_shift=0.001) #md refine(atmsel) cg = conjugate_gradients() cg.optimize(atmsel, max_iterations=200, min_atom_shift=0.001) #molecular dynamics def refine(atmsel): # at T=1000, max_atom_shift for 4fs is cca 0.15 A. md = molecular_dynamics(cap_atom_shift=0.39, md_time_step=4.0, md_return='FINAL') init_vel = True for (its, equil, temps) in ((200, 20, (150.0, 250.0, 400.0, 700.0, 1000.0)), (200, 600, (1000.0, 800.0, 600.0, 500.0, 400.0, 300.0))): for temp in temps: md.optimize(atmsel, init_velocities=init_vel, temperature=temp, max_iterations=its, equilibrate=equil) init_vel = False #use homologs and dihedral library for dihedral angle restraints def make_restraints(mdl1, aln): rsr = mdl1.restraints rsr.clear() s = selection(mdl1) for typ in ('stereo', 'phi-psi_binormal'): rsr.make(s, restraint_type=typ, aln=aln, spline_on_site=True) for typ in ('omega', 'chi1', 'chi2', 'chi3', 'chi4'): rsr.make(s, restraint_type=typ+'_dihedral', spline_range=4.0, spline_dx=0.3, spline_min_points = 5, aln=aln, spline_on_site=True) a = MyModel(env, alnfile=current_file, knowns='3O26', sequence='R1Y', assess_methods=(assess.DOPE, assess.GA341)) a.starting_model = 1 a.ending_model = 1 a.make() So , Could you guide me on single residue optimization on automodel structures ? Thanking you in advance

2 1