- modeller_usage - salilab.org

Re: [modeller_usage] RE : Re: Class mymodel: How to randomize coordinates
by Modeller Caretaker 11 Oct '07

11 Oct '07

Baptiste Legrand wrote: > Thanks for your answer, I didn't know that the default deviation is 4.0. > Is it interessant to set it to 8.0 or more or the models will be too far > from the template? Is there a theorical (estimation) limit to explore > the conformationnal space without deviate totally from the template? The deviation parameter controls the spread of initial models, not the final optimized models - if the optimization is functioning correctly, the optimized models should look like the template anyway. (However, optimization is not a 'solved' problem, which is why the initial model is the sequence threaded onto the template structure rather than, say, the extended chain or randomly placed points.) So in practice changing deviation will not have a large effect on your model generation (unless you set it to 0, in which case no randomization is done). Ben Webb, Modeller Caretaker -- modeller-care(a)salilab.org http://www.salilab.org/modeller/ Modeller mail list: http://salilab.org/mailman/listinfo/modeller_usage

1 0

Class mymodel: How to randomize coordinates
by Baptiste Legrand 10 Oct '07

10 Oct '07

Hi all, I try to make models from a template, with constraints (Helix) and a randomization of the coordinates for each models. I set deviation to 4.0 but MODELLER dont use it. A part of my script: # Add Helix class mymodel(automodel): def special_restraints(self, aln): rsr = self.restraints at = self.atoms rsr.add(secondary_structure.alpha(self.residue_range('2:', '13:'))) # Build models, and assess with both DOPE and GA341 a = mymodel(env, alnfile = 'f125-1s35.ali', # alignment filename (use default: 'a = automodel', 'loopmodel' for loop, 'allhmodel' to add protons) knowns = '1j18', # codes of the templates (for multiple template, just add codes: knowns = ('1s35', '1quu')) sequence = 'f3s5', # code of the target deviation = 4.0, assess_methods =(assess.DOPE, assess.GA341)) # assessment scripts #inifile = 'my-initial.pdb') # use my initial structure What should I do to randomize in this case? Thanks in advance, Regards Baptiste --------------------------------- Ne gardez plus qu'une seule adresse mail ! Copiez vos mails vers Yahoo! Mail

2 1

Re: [modeller_usage] exploding GTP
by Modeller Caretaker 09 Oct '07

09 Oct '07

Jonathan Winger wrote: > I'm using Modeller 9v2. My protein is a homodimer, and my template is > also a homodimer. There is one molecule of AMPCPP and 2 metals bound to > each chain of the template. I adjusted the PDB file (2BW7) so that the > metals and nucleotides followed directly after the appropriate chain > TER's. I should add that the modeling runs fine in the absence of > nucleotide. You could try using the BLK residue type (.) instead of GTP (y) in your model sequence - it is entirely possible that the CHARMM forcefield parameters for GTP are not doing a great job. As for your metal, your model contains MG but your template contains CA, so since you have no equivalent template atom, the MG gets no homology-derived restraints on it and it can wander (Modeller is not smart enough to recognize automatically that one metal could be considered equivalent to another here.) To fix that, you could either create manual distance restraints on your metal ion to hold it in place, or edit your template PDB and change the metal ion there to MG. Ben Webb, Modeller Caretaker -- modeller-care(a)salilab.org http://www.salilab.org/modeller/ Modeller mail list: http://salilab.org/mailman/listinfo/modeller_usage

1 0

exploding GTP
by Jonathan Winger 09 Oct '07

09 Oct '07

When modeling GTP and metal ions into my protein, using a template with 2 metal ions and an ATP analog, the GTP atoms end up disconnected from each other and scattered around the active site, and one of the two metals is launched out of the protein. What can I do to fix this? Thanks much, Jon

2 1

qsub and cis restraints
by Gordon Wells 08 Oct '07

08 Oct '07

Do classes inheriting from automodel get propagated when using the sge_qsub parallel job? I can get a parallel run to work with automodel, but not with a new class including a cis-peptide restraint defined in special_restraints (which I can get to run in single proc mode). I get the following error: # Traceback (most recent call last): File "psamdc.py", line 77, in ? a.make() File "/home/namd/lib/modeller9v2/modlib/modeller/automodel/automodel.py", line 110, in make self.multiple_models(atmsel) File "/home/namd/lib/modeller9v2/modlib/modeller/automodel/automodel.py", line 214, in multiple_models self.parallel_multiple_models(atmsel) File "/home/namd/lib/modeller9v2/modlib/modeller/automodel/automodel.py", line 234, in parallel_multiple_models self.outputs.extend(job.run_all_tasks()) File "/home/namd/lib/modeller9v2/modlib/modeller/parallel/job.py", line 110, in run_all_tasks self.finish_next_task() File "/home/namd/lib/modeller9v2/modlib/modeller/parallel/job.py", line 127, in finish_next_task self.process_event(obj, s) File "/home/namd/lib/modeller9v2/modlib/modeller/parallel/job.py", line 141, in process_event task = obj.task_results() File "/home/namd/lib/modeller9v2/modlib/modeller/parallel/slave.py", line 54, in task_results r = self.get_data(allow_heartbeat=True) File "/home/namd/lib/modeller9v2/modlib/modeller/parallel/communicator.py", line 78, in get_data (cmdtype, obj) = self._recv() File "/home/namd/lib/modeller9v2/modlib/modeller/parallel/communicator.py", line 116, in _recv raise RemoteError(obj, self) modeller.parallel.communicator.RemoteError: AttributeError: 'module' object has no attribute 'cismodel' from <SGE qsub slave, ID 107.1> Is there a convenient way around this? Thanks in advance Regards -- I think perhaps the most important problem is that we are trying to understand the fundamental workings of the universe via a language devised for telling one another when the best fruit is. -- Terry Pratchett, alt.fan.pratchett -- Gordon Wells Bioinformatics and Computational Biology Unit Department of Biochemistry University of Pretoria

2 1

Re: [modeller_usage] Mutations
by laura.buttazzo＠email.it 04 Oct '07

04 Oct '07

Hi, I tried the example in http://salilab.org/modeller/wiki/Mutate_model, but my model hasn't chains. It has been calculated with Modeller and the residue that I would like to change is the 27: ATOM 203 N ASP 27 16.536 9.555 15.986 1.00 91.26 1SG 204 ATOM 204 CA ASP 27 16.891 10.947 16.025 1.00 91.26 1SG 205 ATOM 205 CB ASP 27 18.072 11.218 16.974 1.00 91.26 1SG 206 ATOM 206 CG ASP 27 18.307 12.718 17.052 1.00 91.26 1SG 207 ATOM 207 OD1 ASP 27 17.470 13.485 16.504 1.00 91.26 1SG 208 ATOM 208 OD2 ASP 27 19.335 13.117 17.662 1.00 91.26 1SG 209 ATOM 209 C ASP 27 17.319 11.354 14.653 1.00 91.26 1SG 210 ATOM 210 O ASP 27 17.069 12.473 14.212 1.00 91.26 1SG 211 However I tried to change the following parts in the script: >#first argument >modelname, respos, restyp, chain, = sys.argv[1:] I put: modelname, respos, restyp, = sys.argv[1:] >s = selection(mdl1.chains[chain].residues[respos]) I put s = selection(mdl1.residues[respos]) >s = selection(mdl1.chains[chain].residues[respos]) s = selection(mdl1.residues[respos]) But I have obtained a model very similar as that in input. Should I try to change some parameter in the script? Thank you in advance. Laura --------- Original Message -------- Da: Modeller Caretaker <modeller-care(a)salilab.org> To: Cc: modeller_usage(a)salilab.org Oggetto: Re: [modeller_usage] Mutations Data: 25/09/07 19:46 > > > > laura.buttazzo(a)email.it wrote: > > I'm a new user of Modeller and I have a question about the model computing. > > If I take one/or more models with a given alignment and I change just one or > > more amino acids maintaning the same alignment, should I obtain a different > > model? > > In other words, Can Modeller be used for calculating mutations? > > Certainly - there has been some work done for SNPs using Modeller; see > http://salilab.org/modeller/wiki/Mutate_model for an example script. > > Ben Webb, Modeller Caretaker > -- > modeller-care(a)salilab.org http://www.salilab.org/modeller/ > Modeller mail list: http://salilab.org/mailman/listinfo/modeller_usage > > > -- Email.it, the professional e-mail, gratis per te: http://www.email.it/f Sponsor: In REGALO 'Meravigliosa Creatura' la super hit di GIANNA NANNINI Clicca qui: http://adv.email.it/cgi-bin/foclick.cgi?mid=6615&d=20071004

2 1

entangled domains
by Russell Green 03 Oct '07

03 Oct '07

Hello, I'm new to Modeller. I have a protein segment I'm trying to model that consists of 2 domains. This protein has 4 very similar domains; 2 have been resolved and the other 2 have not (the ones I want to model). I go through the motions of aligning the sequences and building the model; however, the 2 domains I modelled are basically entangled. Thus, my question is, how can I keep this from happening (such as specifying a straight link between the domains instead of the current one that loops back, if this is even plausible)? Any suggestions are greatly appreciated. Thank you, Russell Green

2 1

Models from secondary structure and distance constraints alone?
by Andrew D. Fant 28 Sep '07

28 Sep '07

[it appears that I accidentally sent this to the listowner rather than the list. oops. sorry Ben] This may seem like a rather strange question, but is it possible to get Modeller to output a PDB structure based only on secondary structure information and a few distance constraints (primarily dicystene relationships). I realize that without an alignment to a known structure, anything that is generated probably bears little resemblance to reality, but I have a relatively short fragment that isn't significantly related to anything in the PDB, and between the disulfide linkages and a couple other known constraints, there aren't that many ways that it could bend together. Basically, I'm trying to generate a consistent structure that is in a possible conformation that doesn't violate steric boundaries and such, without having to manually generate the coordinates from the helix and sheet parameters. Can anyone make any suggestions? Thanks, Andy

2 1

Modeling of a C-terminal segment
by Bo Yang 28 Sep '07

28 Sep '07

Hello, Modeller: I have a question of modelling the C-terminal region of my target. The protein matches with the crystal structure templates except the C-terminal region (named as C-tail here). I did BLAST search using this segment to against the proteins with known structures at Protein Data Band. It returns a match with a short region of an antibody at 56% sequence identity with the inclusion of 4 gaps. The function of my target protein has nothing to do with antibody. I am not sure if it is reasonable to use the BLAST match results as templates to model the conformation of the C-tail. I have thought to model this C-tail region using the Loop Model in Modeller. There are some concerns here. First, the C-tail contains 24 amino acids. Will this be too long for the loop-model in Modeller? If the length is not the problem, how reliable are the conformations of the segment? Second, the segment is at the C-terminal of the protein. For my understanding, the Loop Model in Modeller requires the N and C terminus of the segment to be fixed. The position of the last residues of the segment will have significant influence on the Loop conformation. Which one is more reasonable, to model this segment based on the sequence homology, or use a loop modelling algorithm? Your help and inputs will be much appreciated! Bo --------------------------------- Check out the hottest 2008 models today at Yahoo! Autos.

2 1

September 70% OFF
by VIAGRA � Official Site 26 Sep '07

26 Sep '07

1 0