June 2010 - modeller_usage

Fwd: SALIGN
by Piyush Diyora 12 Jun '10

12 Jun '10

> HI, > I performed multiple sequence alignment using salign. When I analysed the result output (.pap output), there was some improper alignment at 400th position. Because of that, entire alignment after that error was improper. How can I improve the alignment and what is the reason for such mistakes. > > I have attached the file containing sequence of the protein, salign code and the pap output file.

3 2

Modeling a dimer
by Konstantin A. Odynets 10 Jun '10

10 Jun '10

Dear Modeller users, my templates are a homodimer of the first domain and a monomer of the second domain, and the target protein is a homodimer. Both domains are linked by 17 aa flexible regions without templates. How could I do such task in Modeller 9.8? >P1;1n3l structureX:1n3l:4:A:342:B:hstyrrs-N_2x342:homo sapiens:1.18:0.18 -AAAAAAAAAA-AAAAAA-------/-BBBBBBBBBB-BBBBBB-------* >P1;1ntg structureX:1ntg:2:A:170:A:hstyrrs-C_1x169:homo sapiens:2.21:0.251 -------------------CCCCCC* >P1;hstyrrs sequence:hstyrrs:1:A:528:B:hstyrrs_2x528:homo sapiens:2.00:-1.00 AAAAAAAAAAAAAAAAAALCCCCCC/BBBBBBBBBBBBBBBBBBLCCCCCC* Thank you in advance for your help. Best regards, Konstantin. --- Konstantin A. Odynets Institute of Molecular Biology and Genetics Academician Zabolotny street, 150, Kyiv, Ukraine E-mail: odynets(a)imbg.org.ua

2 1

Parallel Python (socket.error: (111, 'Connection refused')
by Laurence Cooke 10 Jun '10

10 Jun '10

Hello, I have password-less ssh and rsh (I've tried both; same "Connection refused" error). My home directory is NFS mounted to all slave machines. I successfully ran parallel python with Modeller9v7 in the past with this setup but now since updating to Mod9v8 I'm not able to run the same scripts. It's been a while since I've used Modeller so I'm not sure if the problem is due to updating Modeller or changes that might have occurred in the computer configurations. Again ssh and rsh work seamlessly between all nodes without passwords. Also: I'm able to run the scripts using 'local_slave()'. I updated to Modeller9v8 on all machines. All nodes are correctly listed in the /etc/hosts files on all machines. We do run other software apps that require password-less ssh and they are working well. Nodes are quad core processors. Finally: I included below the code I'm running as well as the Modeller output in the *.slave files. Any help would be greatly appreciated. Regards, Laurence Cooke Arizona Cancer Center Tucson, Arizona ****************************************** Start Parallel Task Remote file (one node) ****************************************** from modeller import * from modeller.parallel import * from buildmodeltask import BuildModelTask log.minimal() j = job() j.append(ssh_slave('slave1', ssh_command='ssh')) j.append(ssh_slave('slave1', ssh_command='ssh')) j.append(ssh_slave('slave1', ssh_command='ssh')) j.append(ssh_slave('slave1', ssh_command='ssh')) j.queue_task(BuildModelTask('alignment_1.ali','pdb','sequence_1')) j.queue_task(BuildModelTask('alignment_2.ali','pdb','sequence_2')) j.queue_task(BuildModelTask('alignment_3.ali','pdb','sequence_3')) j.queue_task(BuildModelTask('alignment_4.ali','pdb','sequence_4')) results = j.run_all_tasks() print "Done: ", results ***************************** BuildModelTask file ****************************** import random from modeller import * from modeller.parallel import task from modeller.automodel import * class BuildModelTask(task): def run(self,ali,pdb,unknown): env = environ(rand_seed=(random.randint(2,50000)*-1)) env.io.hetatm = True a = automodel(env, alnfile=ali, knowns=pdb, sequence=unknown, assess_methods=(assess.DOPE, assess.GA341)) a.starting_model = 1 a.ending_model = 250 a.make() ********************************* *.slave0-4 files ********************************* Slave startup: connect to master at 127.0.0.1:59941:0:LMTRWVLB Traceback (most recent call last): File "/usr/lib/modeller9v8/bin/modslave.py", line 27, in ? slaveloop(sys.argv[2]) File "/usr/lib/python2.4/site-packages/modeller/parallel/slaveloop.py", line 57, in slaveloop master.connect_back(host, port, identifier) File "/usr/lib/python2.4/site-packages/modeller/parallel/slave_communicator.py", line 15, in connect_back s.connect((host, port)) File "<string>", line 1, in connect socket.error: (111, 'Connection refused') DISCLAIMER: This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please notify the system manager. This message contains confidential information and is intended only for the individual named. If you are not the named addressee you should not disseminate, distribute or copy this e-mail.

2 1

how to obtain the results from assess_methods from automodel() object
by Thomas Evangelidis 10 Jun '10

10 Jun '10

Dear Modeller Caretaker, I use assess_methods=(assess.DOPEHR,assess.normalized_dope) within the automodel constructor to evaluate thousands of models during structure prediction. As you see it becomes very difficult to trace these values within the log files, and assign the scores to the correct files and alignments. Is there any easier way to obtain these values from the instance of automodel(), i.e. ... a = MyModel(env, alnfile=alignment, knowns=known_templates, assess_methods=(assess.DOPEHR, assess.normalized_dope), sequence='target') .... a.make() dopehr_score = a.get_dopehr(modelNo) ## something like this thanks in advance, Thomas

2 1

program to manually edit the structure
by Thomas Evangelidis 07 Jun '10

07 Jun '10

Dear Modellers, are you aware of any commercial free software that allows you to flip side-chains manually and correct bad torsion angles?

1 0

Re: [modeller_usage] Running GA341 Assessment
by David Tiberi 03 Jun '10

03 Jun '10

Hi, I have tried to implement this assessment criteria to evaluate some of my homology models. However, after having run all of the scripts, I keep ending up with a value of 0.0000. Attached to this e-mail I have included the outputs I received while trying to run the assessment. Could you please lend some insight on this matter. Thank you in advance, D.T

2 1

Modelling missing residues
by Omar Davulcu 03 Jun '10

03 Jun '10

Hi everyone, I have a crystal structure for an enzyme (1M80) that is missing a loop, residues 311-319. I also have a structure of a homologue in which this loop is not missing, and I was hoping to use this as a template to model the missing loop. I've been using: http://www.salilab.org/modeller/wiki/Missing <http://www.salilab.org/modeller/wiki/Missing%20residues> residues as a guide. Briefly, I generate an initial model of the loop in 1M80 by filling in the sequence of the missing residues and aligning with the sequence in 1M80, inserting gaps for missing residues. This model is then used (with the inifile parameter under the automodel class), along with an alignment of the filled in 1M80 sequence with the homologous structure, to actually homology model the loop. As I don't want the rest of the protein to move, in both of these steps, I'm modifying select_atoms to include only the residues in the loop. The resulting model looks pretty good in terms of the conformation of the loop. What I notice, however, is that the geometry is a little bad where the loop reconnects with the rest of the protein. For example, the peptide bond between residues 310-311 and 319-320 is about 1.55 angstroms (ideal being ~1.33). I've tried "optimizing out" this issue with approaches like: a.library_schedule=autosched.slow a.max_var_iterations=500 a.md_level=refine.very_slow a.repeat_optimization=5 .and even increasing repeat_optimization to 25 cycles. This doesn't seem to help. Also, expanding the residues included in the optimization (from 311-319 to 310-320, for example) just moves the bad geometry to the boundary of that new window. Any ideas on what I'm doing wrong? Or how to fix the geometry within modeller? Thanks for the help! Omar Davulcu

1 0

side-chains like in template
by sdh 01 Jun '10

01 Jun '10

Hello, I want to create a model based on two templates. How can I constraint few selected positions, so they would preserve (if possible) side-chain conformation of one of the templates? Thanks for hints, sdh

2 3