- modeller_usage - salilab.org

different structures
by Gulshat Ibragimova 13 Oct '98

13 Oct '98

Dear modellers: I had got some questions when I tried to build a model on a basis of 3 different structures. I used full_homol.top (or model.top) script of the Modeller4. The sequence identities were 59, 54, and 47%. The resolution of structures was 1.8 - 2 A. The alignment was perfect. Nevertheless, at the end of an optimization procedure, the objective function was ca.3000 - 4000. When I have used only one structure (with the sequence identity 59%), the same optimization procedure resulted in models with the objective function of 600 - 700. Probably, for this particular case only one structure was necessary. However, it is still unclear to me, why using different structures increased the objective function value by an order? Which role does the CLUSTER_CUT option play for the TRANSFER_XYZ command? Changing it to 1.0 (as recommended by Manual) didn't let the Modeller4 run. An error message was reported: trfxyz_E> structures not isomorphic: i,na,na",nr,nr" 2 1147 1129 150 144 What does it mean? Thanks in advance, Gulshat --------------------------------------------------------------------- Gulshat Ibragimova, Ph.D. EMBL, Postfach 102209 69012 Heidelberg, FRG Tel: +49 (0)6221 387466, Fax: +49 (0)6221 387517

3 2

I need help
by A.Romani＠leonardo.biomed.unipr.it 10 Sep '98

10 Sep '98

-------- I'm able to perform the basics of building, but now I've few questions to improve my model. I'built the model with the top script below : _____________________________________________ INCLUDE # Include the predefined TOP routines SET ALNFILE = 'alignment.ali' # alignment filename SET SEQUENCE = 'my_sequence' SET LIBRARY_SCHEDULE = 1 SET MD_LEVEL = 'refine1' SET FINAL_MALIGN3D = 1 SET REPEAT_OPTIMIZATION = 5 SET RSTRS_REFINED = 5 SET MD_RETURN = 'MINIMAL' SET KNOWNS = 'knows1' 'knows2' 'knows3' # codes of the templates SET ATOM_FILES_DIRECTORY = '/disk2/modeller4/examples' # directories for input atom files SET STARTING_MODEL= 1 # index of the first model SET ENDING_MODEL = 20 # index of the last model # (determines how many models to calculate) CALL ROUTINE = 'model' # do homology modelling _____________________________________________________________________________________ With this script am I sure to build an accurate model? Are there other stuff that I can do to improve my model? How can I transfer the heteroatoms from template to model? Thanking in advance Your faithfully Antonello Romani ________________________________ Antonello Romani Istituto di Patologia Generale Plesso Biotecnologico Integrato Universita' degli Studi di Parma via Volturno 39 43100 Parma Tel. +39 521 903 751 Fax. +39 521 903 742 E-mail : A.Romani(a)leonardo.biomed.unipr.it _______________________________

1 0

Re: Modeling a loop
by Robert Jorissen 05 Aug '98

05 Aug '98

Dear fellow MODELLERs I have found that the problem mentioned by Tom Lee about modelling loops only occurs for the MSI version of MODELLER and not the academic version. I get around the problem by putting the following in the TOP file.: INCLUDE DEFINE_STRING VARIABLES = LOOP_CSRFILE LOOP_INI_MODEL SET SEQUENCE = 'aaa' SET MODEL = 'aaa.pdb' # initial model of the target : : The define_string... line is the only line to be added. I can't remember how I trouble-shot the problem as I fixed it a month ago. I believe this will fix Dr Lee's problem. Regards Rob -- ----------------------------------------------------- Robert Jorissen Ludwig Institute for Cancer Research P.O. BOX 2008, Royal Melbourne Hospital Parkville, Victoria 3050, Australia phone: ++61-3-9341-3155 fax: ++61-3-9341-1938 e-mail: Robert.Jorissen(a)ludwig.edu.au -----------------------------------------------------

2 1

Modeling a loop
by Tom Lee 27 Jul '98

27 Jul '98

Hi, I have generated a model for my molecule. I'm trying to study a loop region in detail. I followed the instruction in FAQ of the manual and wrote the following Top file: ------------------------------------------------------------------------------- INCLUDE SET SEQUENCE = 'ASLA2' SET MODEL = 'ASLA2.B99990001.pdb' SET ATOM_FILES_DIRECTORY = './:../atom_files' SET OUTPUT_DIRECTORY = './' SET PDB_EXT = '.pdb' SET STARTING_MODEL = 20 SET ENDING_MODEL = 29 SET MD_LEVEL = 'refine1' SET RAND_SEED = -34871 CALL ROUTINE = 'loop' CALL ROUTINE = 'cluster', ID1 = 20, ID2 = 29 SUBROUTINE ROUTINE = 'select_atoms' PICK_ATOMS SELECTION_SEGMENT = '49:' '76:', SELECTION_STATUS = 'initialize' RETURN END_SUBROUTINE ----------------------------------------------------------------------------- "ASLA2.B99990001.pdb" is the file generated by the previous Modeler run. "49" and "76" are the starting and ending residues of the loop region. I got an error message after running this job: ------------------------------------------------------------------------------- TOP______> 102 414 STRING_OPERATE OPERATION = 'CONCATENATE', ; STRING_ARGUMENTS = SEQUENCE '.mat', RESULT = MATRIX_FI; LE TOP______> 103 415 SET ROOT_NAME = SEQUENCE TOP______> 104 416 RETURN TOP______> 105 591 STRING_OPERATE OPERATION = 'CONCATENATE', ; STRING_ARGUMENTS = '$(JOB).rsr', RESULT = LOOP_CSRFILE assgn___E> cannot assign a value to a value: ------------------------------------------------------------------------------- Do you know what went wrong? Thanks. Tom

2 2

Modeller
by he＠msvax.mssm.edu 27 Jul '98

27 Jul '98

Dear Modellers: I got a problem when i'm modelling my protein. The situation is that: I found my protein (aam) has a partial similarity with 1nkl ('5:' ' 127:'), made a ali. file, and call routine ='model' or using "build_model" . I got a PDB format file or .ini file (for example aam.B999901 or aam.ini). Then, i try to transfer / or modify another segment of aam from a high similarity protein (1udg, '223:' '289:') without losing 1nkl ('5:' '127:') information in aam.ini file, but in vain. Could you tell me which commands fit this purpose? Thanks for help. Xingxuan He, MD. Associate Dept. of Human Genetics Mt. Sinai Medical Center

2 1

molecular modelling
by Dr. Ahsan Salim 27 Jul '98

27 Jul '98

I am trying to model a tetramer. The number of residues in the target protein are 141 each in 2 subunits and 146 each in the other two. The target is single, also a tetramer with the same number of amino acids in the four subunits. Following is the format of the alignment file: C; alignment in the PIR format >P1;----- structureX:-----:1:A:146:D:name:source:: ------------------------------------------------------------------------- ---------------------------------------------/ ------------------------------------------------------------------------- ---------------------------------------------/ ------------------------------------------------------------------------- ---------------------------------------------/ ------------------------------------------------------------------------- ---------------------------------------------* >P1;----- sequence:-----:1:A:146:D:name:source:: ------------------------------------------------------------------------- ---------------------------------------------/ ------------------------------------------------------------------------- ---------------------------------------------/ ------------------------------------------------------------------------- ---------------------------------------------/ ------------------------------------------------------------------------- ---------------------------------------------* Please make sure that the above format is correct. Moreover the total number of residues amounts to above 1000 which is more than the specified limit. Please suggest the solution. Thanks

2 1

modelling
by Prof. Zafar H. Zaidi 27 Jul '98

27 Jul '98

I am trying to model a tetramer. The number of residues in the target protein are 141 each in 2 subunits and 146 each in the other two. The target is single, also a tetramer with the same number of amino acids in the four subunits. Following is the format of the alignment file: C; alignment in the PIR format >P1;----- structureX:-----:1:A:146:D:name:source:: ------------------------------------------------------------------------- ---------------------------------------------/ ------------------------------------------------------------------------- ---------------------------------------------/ ------------------------------------------------------------------------- ---------------------------------------------/ ------------------------------------------------------------------------- ---------------------------------------------* >P1;----- sequence:-----:1:A:146:D:name:source:: ------------------------------------------------------------------------- ---------------------------------------------/ ------------------------------------------------------------------------- ---------------------------------------------/ ------------------------------------------------------------------------- ---------------------------------------------/ ------------------------------------------------------------------------- ---------------------------------------------* Please make sure that the above format is correct. Moreover the total number of residues amounts to above 1000 which is more than the specified limit. Please suggest the solution. Thanks

1 0

Building loops in Modeller
by Robert Jorissen 13 Jul '98

13 Jul '98

Dear fellow Modeller users I am using Modeller 4 to build a comparative model and am at the stage of constructing loops. Do I accept the loop with the lowest score (as long as it is plausible) or do I use other criteria as well e.g. combination of low score and favourable ProsaII plot over the region of the loop. Rob -- ----------------------------------------------------- Robert Jorissen Ludwig Institute for Cancer Research P.O. BOX 2008, Royal Melbourne Hospital Parkville, Victoria 3050, Australia phone: ++61-3-9341-3155 fax: ++61-3-9341-1938 e-mail: Robert.Jorissen(a)ludwig.edu.au -----------------------------------------------------

2 1

precision of compilation
by Jiri Damborsky 09 Jul '98

09 Jul '98

Dear Modeller users; I have been using Modeller v. 3c for some time and more recently started to use v. 4h for my modelling projects. However, version 4 seems to give worse results with regards to stereochemical quality (e.g., checked by PROCHECK). I used the same top.file for both versions but the results were different. Very recently I found out, that my v. 4h is compiled with single precision, while v. 3c is compiled with double precision. Could this be the reason for difference in quality of results? Is the version 4 available with double precision? Thank you for your reply. Best wishes, Jiri * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * Dr. Jiri Damborsky Laboratory of Biomolecular Structure & Dynamics Masaryk University Kotlarska 2 611 37 Brno Czech Republic e-mail: jiri(a)chemi.muni.cz phone: +420-5-41129 377 FAX: +420-5-41129 506 http://web.chemi.muni.cz/~jiri

1 0

Re: Creating new residue types
by Andrej Sali 09 Jul '98

09 Jul '98

Another note on new residue types: Version 5 of modeller, to be released by the end of August, is compatible with the QUANTA RTF editor. So the easiest way to generate a new RTF entry would be to use QUANTA's RTF editor, write it out, and include it in the MODELLER lib file. Also, we will have many more molecules/residues in the standard MODELLER RTF library to begin with. I know, not everybody has access to QUANTA ... Best, Andrej -- Andrej Sali, Assistant Professor The Rockefeller University, 1230 York Avenue, New York, NY 10021-6399 tel +1 212 327 7550; lab +1 212 327 7206 ; fax +1 212 327 7540 e-mail sali(a)rockvax.rockefeller.edu; http://salilab.org

2 1