- modeller_usage - salilab.org

Fwd: how to ERROR: MODEL atom not classified: TYR:OT1 TYR
by Modeller Caretaker 10 Sep '03

10 Sep '03

Forwarding to list. ----- Forwarded message from Sherin Yang <sherin(a)sina.com> ----- From: "Sherin Yang" <sherin(a)sina.com> To: <modeller-care(a)salilab.org> Subject: how to ERROR: MODEL atom not classified: TYR:OT1 TYR Date: Wed, 10 Sep 2003 10:54:04 +0800 X-Mailer: Microsoft Outlook Express 6.00.2800.1106 Hi, all I want to use Align2D to align a sequence with a structure with pdb format. I can't get the proper result. The log file tell me the following wrong: iatmcls_286W> MODEL atom not classified: TYR:OT1 TYR iatmcls_286W> MODEL atom not classified: TYR:OT2 TYR Who can tell me how to resolve such problem? Thank you! Sherin Yang ----- End forwarded message -----

1 0

Re: model_refinement
by Modeller Caretaker 08 Sep '03

08 Sep '03

Forwarding to list. ----- Forwarded message from Denis Volkov <themelon(a)yandex.ru> ----- Date: Mon, 8 Sep 2003 17:41:54 +0400 From: Denis Volkov <themelon(a)yandex.ru> To: Modeller Caretaker <modeller-care(a)salilab.org> Subject: Re: Fwd: Re: model_refinement Hello Modeller, Thursday, September 04, 2003, 12:16:20 AM, you wrote: MC> ----- Forwarded message from Oliver Hucke <ohucke(a)u.washington.edu> ----- MC> Date: Tue, 02 Sep 2003 10:03:00 -0700 MC> From: Oliver Hucke <ohucke(a)u.washington.edu> MC> To: Modeller Care <modeller-care(a)salilab.org> MC> Subject: Re: model_refinement MC> Hi Denis, MC> did you compare the quality of the structure calculated with modeller MC> with that of the template(s)? MC> If the quality of the model is really worse something might be wrong MC> with your alignment. Did you do the manual adjustment of the input MC> alignment on basis of the analysis of the quality of the first models? MC> This usually helps a lot. MC> I do not think that it is necessary to use charmm or another external MC> molecular mechanics package to improve the quality of your model. MC> Modeller actually uses (an older version of) the charmm force field for MC> the molecular mechanics part of the optimization. MC> Best, MC> Oliver MC> Modeller Care wrote: >>------ Forwarded Message >> >>From: Denis Volkov <D.Volkov(a)artcustoms.ru> >>X-Mailer: The Bat! (v1.52f) >>Reply-To: Denis Volkov <D.Volkov(a)artcustoms.ru> >>X-Priority: 3 (Normal) >>Message-ID: <43123585810.20030902125313(a)artcustoms.ru> >>To: modeller <modeller_usage(a)salilab.org> >>Subject: model_refinment >> >> >>Hello everyone, >>I successfully used Modeller to model unknown protein. Now I'd like >>to sent my model to Protein Data Bank. So i checked my structure >>using Procheck and find some problems, namely some bad contacts,etc. >>In order to clean a structure i used Charmm program for energy >>minimization. But after it i obtain structure with bad angles and >>bond lengths. Can anybody know how to clean a structure well for >>PDBank and what programs should be use for it. >>Thanks in advance >>Denis Volkov >> >>Enzyme Laboratory,IBCH RAS >> >> >> >> Hello,Oliver Thank you for fast answering! :) No i didn't test a quality of templates course i used 4 structures as a templates so its difficult to do it. Alignment I manually adjust so it's not a origin of fault. I look at the model again and think it's not so bad as I said. But a problem is more global: A similarity with templates is no more than 20%. A great deal of mismatches are in loop origin so it's not so bad...but can i do a good model with so low similarity? As I said I use Charmm, I used it to investigate protein-protein interactions. I used Modeller model to investigate enzyme-substrate and enzyme-inhibitor interaction. But after MD a structure was a great deal disturbed. Is it a problem of low similarity or as you said its not good to use MD with models maid by Modeller? Also I've another question: earlier I used Windows version of Modeller and everything was OK, but now I run Modeller on Linux and have some problem. When I used sequence-search routine it runs for some time (~30-40 min) but then terminated. In log file there is nothing interesting: chains libraries were opened two times. First times it takes 20Mb of memory, second time ~200 Mb and then it terminated. So the problem is a memory usage as I think. I looked through archive and find out that it's a common problem. I checked my top-files and system config (stack size) and everything is good. I've got 500 Mb RAM but half of it utilized by X-Windows. So may be i should run some Modeller routines without running X ? Thanks in advance Denis Volkov Enzyme Laboratory,IBCH RAS -- Best regards, Denis mailto:themelon@yandex.ru ----- End forwarded message -----

1 0

Fwd: can't write the model structure
by Modeller Caretaker 05 Sep '03

05 Sep '03

Forwarding to list. ----- Forwarded message from Sherin Yang <sherin(a)sina.com> ----- From: "Sherin Yang" <sherin(a)sina.com> To: <modeller-care(a)salilab.org> Subject: can't write the model structure Date: Sat, 6 Sep 2003 02:17:46 +0800 X-Mailer: Microsoft Outlook Express 6.00.2800.1106 Hi, all I have write a TOP script and want to align a sequence with a structure using ALIGN2D. but I can't get the proper alignment result. The structure's sequence isn't contained in the alignment file. And the log file didn't inform any error information. The following is the TOP scrip, and the attachment is the PDB file. The 1CF8 contains two chains, L and H. Maybe there is something wrong in "MODEL_SEGMENT='1:L' 'X:L' "? ######################################################## READ_MODEL FILE = '1CF8.pdb', MODEL_SEGMENT='1:L' 'X:L' SEQUENCE_TO_ALI ALIGN_CODES = '1CF8L' READ_ALIGNMENT FILE = query_mod.ali', ALIGN_CODES = 'query' ALIGN2D WRITE_ALIGNMENT FILE = 'align2d.ali', ALIGNMENT_FORMAT = 'PIR' ########################################################### Thank you ! Sherin ----- End forwarded message -----

1 0

Fwd: Re: model_refinement
by Modeller Caretaker 03 Sep '03

03 Sep '03

----- Forwarded message from Oliver Hucke <ohucke(a)u.washington.edu> ----- Date: Tue, 02 Sep 2003 10:03:00 -0700 From: Oliver Hucke <ohucke(a)u.washington.edu> To: Modeller Care <modeller-care(a)salilab.org> Subject: Re: model_refinement Hi Denis, did you compare the quality of the structure calculated with modeller with that of the template(s)? If the quality of the model is really worse something might be wrong with your alignment. Did you do the manual adjustment of the input alignment on basis of the analysis of the quality of the first models? This usually helps a lot. I do not think that it is necessary to use charmm or another external molecular mechanics package to improve the quality of your model. Modeller actually uses (an older version of) the charmm force field for the molecular mechanics part of the optimization. Best, Oliver Modeller Care wrote: >------ Forwarded Message > >From: Denis Volkov <D.Volkov(a)artcustoms.ru> >X-Mailer: The Bat! (v1.52f) >Reply-To: Denis Volkov <D.Volkov(a)artcustoms.ru> >X-Priority: 3 (Normal) >Message-ID: <43123585810.20030902125313(a)artcustoms.ru> >To: modeller <modeller_usage(a)salilab.org> >Subject: model_refinment > > >Hello everyone, >I successfully used Modeller to model unknown protein. Now I'd like >to sent my model to Protein Data Bank. So i checked my structure >using Procheck and find some problems, namely some bad contacts,etc. >In order to clean a structure i used Charmm program for energy >minimization. But after it i obtain structure with bad angles and >bond lengths. Can anybody know how to clean a structure well for >PDBank and what programs should be use for it. >Thanks in advance >Denis Volkov > >Enzyme Laboratory,IBCH RAS > > > > -- ____________________________________________________________ Oliver Hucke (PostDoc) Biomolecular Structure Center Dept. of Biochemistry University of Washington Health Sciences Building K-418C 1959 NE Pacific Ave Seattle, WA 98195 T:(206)685-7046 _______________________________ ----- End forwarded message -----

1 0

modeling membrane protein loops
by Modeller Care 02 Sep '03

02 Sep '03

------ Forwarded Message From: Lucy Forrest <lrf2103(a)columbia.edu> Subject: modeling membrane protein loops Hi, In my experience (see ref below) the membrane does not affect modeling loops to any great extent. However, when comparing different loop conformations energetically, if you treat the loop environment as vacuum then will get a much worse prediction than if you use solvation energy terms also. Discrimination of native loop conformations in membrane proteins: Decoy library design and evaluation of effective energy scoring functions Proteins: Structure, Function, and Genetics Volume 52, Issue 4, Date: September 2003, Pages: 492-509 Lucy R. Forrest, Thomas B. Woolf I can discuss this in more detail if anybody wishes, Lucy Lucy Forrest Honig Lab Department of biochemistry Columbia University New York, NY 10032 >Date: 27 Aug 2003 11:37:52 -0400 >From: Andras Fiser <afiser(a)aecom.yu.edu> >Subject: Re: loop modeling > >Hello, > >It is right. Or as good as it gets, for two reasons, because Modeller >does not account for the environment explicitly anyway and loop modeling >will effect only the loop regions while keeping the rest of the >structure intact, irrespectively of how properly the rest was built. > >You will find a web interface to loop modeling in Modeller at >http://alto.compbio.ucsf.edu/modloop//modloop.html >that may help. > > >best wishes,Andras > >On Wed, 2003-08-27 at 02:32, Moshe Amitay wrote: > > >>Hi all, >>I whuld like your advice: >>I build a model of a membrane protein and I want now to do "loop modeling" >>to few extra cellular and intra-cellular loops. >> I wonder if this protocol of modeller is the right >>choice since the model doesn't contain the membrane and I actually ignore >>the influence of the polar heads. >> thank's >> Moshe Amity. >> >> >> >> >> ------ End of Forwarded Message

1 0

model_refinement
by Modeller Care 02 Sep '03

02 Sep '03

------ Forwarded Message From: Denis Volkov <D.Volkov(a)artcustoms.ru> X-Mailer: The Bat! (v1.52f) Reply-To: Denis Volkov <D.Volkov(a)artcustoms.ru> X-Priority: 3 (Normal) Message-ID: <43123585810.20030902125313(a)artcustoms.ru> To: modeller <modeller_usage(a)salilab.org> Subject: model_refinment Hello everyone, I successfully used Modeller to model unknown protein. Now I'd like to sent my model to Protein Data Bank. So i checked my structure using Procheck and find some problems, namely some bad contacts,etc. In order to clean a structure i used Charmm program for energy minimization. But after it i obtain structure with bad angles and bond lengths. Can anybody know how to clean a structure well for PDBank and what programs should be use for it. Thanks in advance Denis Volkov Enzyme Laboratory,IBCH RAS -- Best regards, Denis mailto:D.Volkov@artcustoms.ru ------ End of Forwarded Message

1 0

FW: building a new residue
by Modeller Care 28 Aug '03

28 Aug '03

------ Forwarded Message From: Tom Burghardt <burghardt(a)mayo.edu> Date: Thu, 28 Aug 2003 14:40:36 -0500 (CDT) To: Modeller Care <modeller-care(a)salilab.org> Subject: building a new residue Dear Modellers, I defined a new residue type following the instructions (a-d) of FAQ17 in the Modeller handbook. I did not do the last step (e) because I need to run the program first to see what needs to be done. To allow Modeller to recognize the new residue single letter code in the sequence file (FAQ17 part d) I modified the residue type library 'modlib/restyp.lib' with a new entry at the end of the file (index=142). When I run Modeller I get the message: read_al_375E> Unknown residue type,position,sequence: U 84 1 recover____E> ERROR_STATUS >= STOP_ON_ERROR: 1 1 implying I guess that it does not recognize the one letter code I assigned my modified residue. Everything runs OK if I replace the modified residue with a BLK residue in the alignment file. My top file is below. INCLUDE # Include the predefined TOP routines SET OUTPUT_CONTROL = 1 1 1 1 1 # uncomment to produce a large log SET ALNFILE = 'seq-sk_sk_T1_short_gly.ali' # alignment filename SET KNOWNS = '2tnx' # codes of the templates SET SEQUENCE = '2tns' # code of the target SET TOPOLOGY_MODEL = 3,HETATM_IO=ON SET TOPLIB = '${LIB}/top_heav.lib' SET PARLIB = '${LIB}/par.lib' READ_RESTYP_LIB RESTYP_LIB_FILE='${LIB}/restyp.lib' SET ATOM_FILES_DIRECTORY = './' # directories for input atom files SET STARTING_MODEL= 1 # index of the first loop model SET ENDING_MODEL = 1 # index of the last loop model CALL ROUTINE = 'model' # do homology modelling Thanks Tom Burghardt ------ End of Forwarded Message

1 0

More weight for user distance restraints
by Modeller Care 28 Aug '03

28 Aug '03

Forwarded by list-owner ------ Forwarded Message From: "Thorsten Schweikardt" <thorsten(a)mail.uni-mainz.de> To: <modeller_usage(a)salilab.org> Subject: More weight for user distance restraints Date: Thu, 28 Aug 2003 16:28:34 +0200 Hi to all, I have a problem with user defined distance restraints, I want to model the n-terminus of a protein that's available as pdb with the help of measured distances. So far, it works, but one little problem remains: the distance-requirements are not met entirely, because modeller does not distort the rest of the protein at all, is there a possibility to force it to add more weight to these distance restraints, even if it means distorting the template protein? Thanks, Thorsten University of Mainz My Top-File: ----------------------------- INCLUDE SET ALNFILE = 'alignment.ali' SET KNOWNS = 'protein1' SET SEQUENCE = 'protein1plusnterm' SET ATOM_FILES_DIRECTORY = './' SET STARTING_MODEL= 1 SET ENDING_MODEL = 20 SET LIBRARY_SCHEDULE = 4 SET FINAL_MALIGN3D = 1 CALL ROUTINE = 'model' SUBROUTINE ROUTINE = 'special_restraints' SET ATOM_IDS = 'OG:3' 'OG:169' ADD_RESTRAINT RESTRAINT_PARAMETERS = 3 1 1 27 2 2 0 30.0 10.0 SET ATOM_IDS = 'OG:116' 'OG:169' ADD_RESTRAINT RESTRAINT_PARAMETERS = 3 1 1 27 2 2 0 40.0 0.0 SET ATOM_IDS = 'CA:42' 'CA:169' ADD_RESTRAINT RESTRAINT_PARAMETERS = 3 1 1 27 2 2 0 40.0 1.0 SET ATOM_IDS = 'CA:3' 'CA:42' ADD_RESTRAINT RESTRAINT_PARAMETERS = 3 1 1 27 2 2 0 21.0 4.0 SET ATOM_IDS = 'CA:3' 'CA:116' ADD_RESTRAINT RESTRAINT_PARAMETERS = 3 1 1 27 2 2 0 50.0 1.0 RETURN END_SUBROUTINE ------ End of Forwarded Message

1 0

loop modeling
by Moshe Amitay 27 Aug '03

27 Aug '03

Hi all, I whuld like your advice: I build a model of a membrane protein and I want now to do "loop modeling" to few extra cellular and intra-cellular loops. I wonder if this protocol of modeller is the right choice since the model doesn't contain the membrane and I actually ignore the influence of the polar heads. thank's Moshe Amity. -------------------------------------- Moshe Amity amitaym(a)mail.biu.ac.il Dep. of Chemistry Bar Ilan University Israel.

2 1

Multidomain protein alignment
by Morgane Agez 27 Aug '03

27 Aug '03

Hi, I have seen in the Chapter 4 of the manual how to build a multi-domain protein model using templates corresponding only to the individual domain. But I think I obtain a model with an incorrect relative orientation of the diferent domains. How can I orient the three template structures appropriatly before the modeling? Thank you very much! With best regards, Morgane

2 1