- modeller_usage - salilab.org

Subject: models for MR
by Bozidar Yerkovich 24 Sep '02

24 Sep '02

>From Katherine Kantardjieff: We have tried this with some success on several TB proteins as part of = the efforts of the TB Structural Genomics Consortium. These have not yet = been formally published. Basically, we have produced both homology and = comparative models using SwissModeller or Modeller, as well as other = servers such as 3D PSSM and the prediction servers at LLNL. Each model = is slightly different, but can be evaluated using whatCheck, ProCheck, = ProsaII etc. We have also used different MR algorithms, and epmr and = amore seem to work the best, although one may find a solution while the = other won't. My advice is to try each model and each MR algorithm = available. We then take the MR results and run the initial maps through phase bias = removal shake& warp (Segelke and Rupp)before viewing. In the best case, = we had an interpretable map and model that could be rebuilt in a day = with XtalView. In the worst case, we obtained a map with interpretable = features, but still poor connectivity and a great deal of work ahead. In = this case, the hybrid fold and sequence match was low to anything in the = available databases. But, we may still be able to pull it out. Professor Katherine Kantardjieff Director, W.M. Keck Foundation Center for Molecular Structure Department of Chemistry and Biochemistry California State University Fullerton 800 N. State College Blvd. Fullerton, CA 92834-6866 Office: (714) 278-3752 eFax: (734) 939-4225 eMail: <mailto:kkantardjieff@fullerton.edu> kkantardjieff(a)fullerton.edu =20

1 0

/pipe1/databases/pdb/uncompressed_files/pdb1hci.ent
by Bozidar Yerkovich 24 Sep '02

24 Sep '02

Hi all, I noticed that some people reported the following error: /pipe1/databases/pdb/uncompressed_files/pdb1hci.ent Command line: INCLUDE assgn___525E> Error in TOP variable assignment. recover____E> ERROR_STATUS >= STOP_ON_ERROR: 1 1 To fix this go to modlib directory where you installed MODELLER and remove pdbent.lib file. Good luck, Bozidar -- Bozidar Yerkovich, PhD Postdoctoral Research Associate Sali Lab, Rockefeller University New York, NY, 10021, USA 212-327-7206 bozidar(a)salilab.org http://salilab.org/modeller

1 0

HETATM_IO
by Ralf Schmid 24 Sep '02

24 Sep '02

Dear all, I'm trying to model a protein including a cofactor, but get an error message pointing towards problems in the alignment and/or pdb file. Modelling without cofactor (i.e. removing "/." from the alignment file) works fine. I can't figure out what causes the problem. Any help will be appreciated, please see the files below. Thanks in advance, ralf xxx.log: <snip> TOP_________> 120 106 READ_ALIGNMENT FILE = ALNFILE, ALIGN_CODES = KNOWNS SEQUE; NCE Dynamically allocated memory at amaxseq [B,kB,MB]: 2205269 2153.583 2.103 openf5__224_> Open 11 OLD SEQUENTIAL xxx.ali openf5__224_> Open 13 OLD SEQUENTIAL ./yyy.pdb Dynamically allocated memory at amaxbnd [B,kB,MB]: 9440821 9219.552 9.003 openf5__224_> Open 11 OLD SEQUENTIAL xxx.ali openf5__224_> Open 13 OLD SEQUENTIAL ./yyy.pdb rdpir___648E> Alignment sequence not found in PDB file: 1 ./yyy.pdb recover____E> ERROR_STATUS >= STOP_ON_ERROR: 1 1 xxx.top: INCLUDE SET OUTPUT_CONTROL= 1 1 1 1 1 SET ALNFILE = 'xxx.ali' SET KNOWNS = 'yyy' SET SEQUENCE = 'xxx' SET STARTING_MODEL = 20 SET ENDING_MODEL = 20 SET HETATM_IO = on CALL ROUTINE = 'model' xxx.ali: >P1;xxx sequence:xxx:.:.:.:.:.:.:.:. VKNER/.* >P1;yyy structureX:yyy:.:.:.:.:.:.:.:. VKQER/.* yyy.pdb: <snip> ATOM 3488 OXT ARG A 447 31.403 66.404 103.394 1.00 65.84 O TER 3489 ARG A 447 HETATM 3490 AP FAD 448 60.780 56.948 67.003 1.00 33.65 P <snip> ----------------------------------------------------------------- Dr. Ralf Schmid ICMB The University of Edinburgh Michael Swann Building 3.14 -----------------------------------------------------------------

3 2

Problem with --error --- SEQUENCE_SEARCH --
by Juan Palacios 24 Sep '02

24 Sep '02

My .top file is: -------------------------------------------------------------------------------------------------------------- INCLUDE SET OUTPUT_CONTROL = 1 1 1 1 1 SET SEARCH_RANDOMIZATIONS = 3 SET FILE = 'TvLDH.ali' SEQUENCE_SEARCH ALIGN_CODES = 'TvLDH', DATA_FILE = ON ----------------------------------------------------------------------------------------------------------------- the log file is: ---------------------------------------------------------------------------------------------------------------- MODELLER 6v2, 17 Feb 2002 PROTEIN STRUCTURE MODELLING BY SATISFACTION OF SPATIAL RESTRAINTS Copyright(c) 1989-2002 Andrej Sali All Rights Reserved Written by A. Sali with help from A. Fiser, R. Sanchez, M.A. Marti-Renom, B. Jerkovic, A. Badretdinov, F. Melo, J.P. Overington & E. Feyfant Rockefeller University, New York, USA Harvard University, Cambridge, USA Imperial Cancer Research Fund, London, UK Birkbeck College, University of London, London, UK Kind, OS, HostName, Kernel, Processor: 4, Linux bioinf4 2.4.4-4GB i686 Date and time of compilation : 05/24/2002 09:27:23 Job starting time (YY/MM/DD HH:MM:SS): 2002/09/23 15:16:01.528 TOP_________> 105 705 SET SEARCH_RANDOMIZATIONS = 3 TOP_________> 106 706 SET FILE = 'TvLDH.ali' TOP_________> 107 707 SEQUENCE_SEARCH ALIGN_CODES = 'TvLDH', DATA_FILE = ON openf5__224_> Open 11 OLD SEQUENTIAL ${MODINSTALL6v2}/modlib//CHAINS_3.0_40_XN.cod Dynamically allocated memory at amaxchn [B,kB,MB]: 2306323 2252.269 2.199 openf5__224_> Open 11 OLD SEQUENTIAL ${MODINSTALL6v2}/modlib//CHAINS_3.0_40_XN.cod openf5__224_> Open 11 OLD SEQUENTIAL ${MODINSTALL6v2}/modlib//CHAINS_all.seq Dynamically allocated memory at amresdb [B,kB,MB]: 5689471 5556.124 5.426 openf5__224_> Open 11 OLD SEQUENTIAL ${MODINSTALL6v2}/modlib//CHAINS_all.seq Dynamically allocated memory at amaxseq [B,kB,MB]: 5877209 5739.462 5.605 Dynamically allocated memory at amaxbnd [B,kB,MB]: 110299725 107714.600 105.190 openf5__224_> Open 11 OLD SEQUENTIAL ${MODINSTALL6v2}/modlib//as1.sim.mat rdrrwgh_268_> Number of residue types: 20 openf5__224_> Open 11 OLD SEQUENTIAL TvLDH.ali --------------------------------------------------------------------------------------------------------------------------------------------------------------- After 2 hours of executing delivery the following mistake The error is: /usr/local/modeller6v2/bin/mod6v2 : line38 : 2030 terminated $NICE ${EXECUTABLE} $STEERF >$LOGF WHAT IS A PROBLEM???

1 0

compare problems
by Alex Brown 24 Sep '02

24 Sep '02

Hi, again. Thanks to all those who answered my previous questions. I haven't had time to checkout the SEQUENCE_SEARCH suggestion, but if MODELLER takes 2 hours or more to find a suitable template, perhaps a BLAST search of the PDB or ExPDB databases would be quicker. Also, the MODELLER ftp server is up and running, and I have downloaded and looked at some of the files. In particular, I have downloaded the structure files associated with the TvLDH example. This leads me on to my next problem - I have downloaded the file 'compare.top' and the structure files 1bmdA.pdb, 4mdhA.pdb, 5mdhA.pdb and 7mdhA.pdb, along with TvLDH.ali. Running the 'compare.top' script produced the following (edited) .log file: Multiple dynamic programming alignment (MALIGN): Residue-residue metric : ${MODINSTALL6v2}/modlib//as1.sim.mat ALIGN_BLOCK : 1 Gap introduction penalty: -900.0000 Gap extension penalty : -50.0000 Length of alignment : 355 pdbnam__217W> Filename for PDB code not found: 1bmd Directories: ./ Extensions : :.atm:.pdb:.ent:.crd rdabrk__288W> Protein not accepted: 1 pdbnam__217W> Filename for PDB code not found: 4mdh Directories: ./ Extensions : :.atm:.pdb:.ent:.crd rdabrk__288W> Protein not accepted: 2 pdbnam__217W> Filename for PDB code not found: 5mdh Directories: ./ Extensions : :.atm:.pdb:.ent:.crd rdabrk__288W> Protein not accepted: 3 pdbnam__217W> Filename for PDB code not found: 7mdh Directories: ./ Extensions : :.atm:.pdb:.ent:.crd rdabrk__288W> Protein not accepted: 4 fndatmi_284E> Beginning atom index for a residue out of bounds: 0 0 recover____E> ERROR_STATUS >= STOP_ON_ERROR: 1 1 Dynamically allocated memory at finish [B,kB,MB]: 14267969 13933.563 13.607 Starting time : 2002/09/24 11:50:17.110 Closing time : 2002/09/24 11:50:45.289 Total CPU time [seconds] : 12.99 I edited the compare.top file, putting an INCLUDE command as the first line. This produced the following (again, edited) .log file: Multiple dynamic programming alignment (MALIGN): Residue-residue metric : ${MODINSTALL6v2}/modlib//as1.sim.mat ALIGN_BLOCK : 1 Gap introduction penalty: -900.0000 Gap extension penalty : -50.0000 Length of alignment : 355 fndatmi_284E> Beginning atom index for a residue out of bounds: 0 0 recover____E> ERROR_STATUS >= STOP_ON_ERROR: 1 1 Dynamically allocated memory at finish [B,kB,MB]: 14267969 13933.563 13.607 Starting time : 2002/09/24 11:53:44.067 Closing time : 2002/09/24 11:53:59.605 Total CPU time [seconds] : 13.67 Could someone please where the error is, and why, initially, the .pdb files are not seen. I am running MODELLER on Mac OSX (Powerbook G4). Thanks, Alex.

1 1

SEQUENCE_SEARCH problems
by Alex Brown 23 Sep '02

23 Sep '02

Hi. I recently downloaded and installed MODELLER 6v2 onto a Macintosh PowerBook G4 under OS X and Darwin. The installation was successful and running the tutorial from the manual (section 1.8.2) produced a result. I am now having a look at the examples provided by the manuscripts on the User Manual page of the Modeller web-site. At this point, I must point out that I have been unable to download the example files - I keep getting a connection failure error. I have started by going through the TvLDH (Trichomonas vaginalis lactate dehydrogenase) example. I created the file 'TvLDH.ali' as shown in the manuscipt, and created the seq_search.top file aslo as detailed : SET SEARCH_RANDOMIZATIONS = 100 SEQUENCE_SEARCH FILE = 'TvLDH.ali', ALIGN_CODES = 'TvLDH', DATA_FILE = ON I run this by 'mod seq_search.top'. However, the program appears just to sit there not doing much - 30 to 40 minutes so far, before I have to quit by Ctrl-C. A .log file is produced, but only contains the introductory title and execution start time. is this normal behaviour. Changing the SEARCH_RANDOMIZATIONS to 20 appears to have no effect. As a supplementary question, does anyone know where I could download the PDB (or even the ExPDB) database - sequences only, in a FASTA or BLAST format ? Cheers, Alex Brown.

3 2

Re: SEQUENCE_SEARCH problems
by Bozidar Yerkovich 23 Sep '02

23 Sep '02

Alex, 1. MODELLER ftp services are down indeed. I apologize for that. We had some security problems, but they have been solved. Those files should be accessible again by the end of a day. 2. SEQUENCE_SEARCH can indeed take a long time. It depends on your machine's CPU and RAM. Just let it run, maybe in the evening. If it's still running next morning that something is wrong. It should finish in 2 hours or so, but then again this might be way off for your particular computer. 3. I am sure if you try google, you find them somehow. God luck, Bozidar > Hi. > > I recently downloaded and installed MODELLER 6v2 onto a Macintosh > PowerBook G4 under OS X and Darwin. The installation was successful and > running the tutorial from the manual (section 1.8.2) produced a result. > > I am now having a look at the examples provided by the manuscripts on > the User Manual page of the Modeller web-site. At this point, I must > point out that I have been unable to download the example files - I keep > getting a connection failure error. > > I have started by going through the TvLDH (Trichomonas vaginalis lactate > dehydrogenase) example. I created the file 'TvLDH.ali' as shown in the > manuscipt, and created the seq_search.top file aslo as detailed : > > SET SEARCH_RANDOMIZATIONS = 100 > SEQUENCE_SEARCH FILE = 'TvLDH.ali', ALIGN_CODES = 'TvLDH', DATA_FILE = ON > > I run this by 'mod seq_search.top'. However, the program appears just to > sit there not doing much - 30 to 40 minutes so far, before I have to > quit by Ctrl-C. A .log file is produced, but only contains the > introductory title and execution start time. is this normal behaviour. > Changing the SEARCH_RANDOMIZATIONS to 20 appears to have no effect. > > As a supplementary question, does anyone know where I could download the > PDB (or even the ExPDB) database - sequences only, in a FASTA or BLAST > format ? > > Cheers, > > Alex Brown. -- Bozidar Yerkovich, PhD Postdoctoral Research Associate Sali Lab, Rockefeller University New York, NY, 10021, USA 212-327-7206 bozidar(a)salilab.org http://salilab.org/modeller ------ End of Forwarded Message

2 1

HETATM_IO and strange log
by Frank J A B 23 Sep '02

23 Sep '02

Dear All I'm facing two problems with modeller. In the FAQ's question #16 there is an explanation on how to use hetero atoms in modelling adding the line HETATM_IO = on in the top script. There is also a comment (not clear to me) about the use of Ws for water and .s for other atoms in the alignment that still remains unsolved to me. In my case I'm trying to use FAD during the modelling. I can't wonder how to edit the alignment to include this Ws and .s I understand that the backslash is a sequence break, but I deduce (not sure) that you must to repeat for each residue a symbol like w, . or whatever in the same .pir file after the backslash. I didn't find any explanation along the manual regard this. Check the example given. INCLUDE SET ALIGNMENT_FILE = 'align1.ali', KNOWNS='templ1', SEQUENCE='targ1' SET HETATM_IO = on, WATER_IO = on CALL ROUTINE = 'model' The alignment file: >P1;templ1 structureX:templ1:1::10:: FAYVI/.wwww* >P1;targ1 sequence:targ1:1::8:: -GWIV/.ww-w* Need I to repeat the entire alignment (template/target) inside the same .pir file (but now using only signs like - , ., w , #) where the .s/Ws positions correpond to those aa that make contact with the HETATM in the template ? The second question is about my log file. I'm habituated to use regularly Modeller since version 4 but suddenly the job was finished abruptly with this : MODELLER 6v2, 17 Feb 2002 PROTEIN STRUCTURE MODELLING BY SATISFACTION OF SPATIAL RESTRAINTS assgn___544E> Variable name not recognized: /pipe1/databases/pdb/uncompressed_files/pdb1hci.ent Command line: INCLUDE assgn___525E> Error in TOP variable assignment. recover____E> ERROR_STATUS >= STOP_ON_ERROR: 1 1 Is this some kind of bug or a user error ? I'm sure that the files are typed correctly (case-sensitive) and the folder where the job is started has all the files in it (the tops, ali and pdb files). I'm not using any file located in the path shown above : /pipe1/databases/pdb/uncompressed_files/pdb1hci.ent Any help will be welcome Thanks in advance Frank Alarcon LNCC - Brasil www.lncc.br

2 1

Bounced emails
by Bozidar Yerkovich 20 Sep '02

20 Sep '02

Hi everybody, Lately I have noticed an increase in the number of bounced emails. Usually you will receive an automated email message saying: Hi, Your submission to the list was bounce because: 1. You are not subscribed to the list, or 2. You are subscribed under a different email address, or 3. You used the word that triggered security mechanism that prevents the list from spam emails. If you wish to resubmit your post, please either 1. Send an email form the same address you used to subscribe yourself, or 2. Subscribe yourself to the modeller_usage list following the instructions from the MODELLER web site, or 3. Re-submit the post omitting the offending word (it's usually help, subscribe, invest, etc.) If you don't remember under which email address you are subscribed, feel free to drop me an email, and I'll check it for you. Or you can simply re-subscribe yourself again. Either way, please don't be mad at me for implementing all this security. Modeller_usage list gets about 10 or so spam emails every day. Without filtering, all of them would end up in your mail-box. Happy modeling, Bozidar Bozidar Yerkovich, PhD Postdoctoral Research Associate Sali Lab, Rockefeller University New York, NY, 10021, USA 212-327-7206 bozidar(a)salilab.org http://salilab.org

1 0

use of Modeller in molecular replacement (X-ray cristallography)
by Karsten Suhre 19 Sep '02

19 Sep '02

Dear Modeller-users, presently I am trying to use Modeller for molecular replacement (determination of protein structure by X-ray cristallography). I would be greately interested in getting into contact with people who have already some experience with this, or learn about published work on this domain. Thank you very much for any hint, Karsten Suhre. -- Charge de Recherche CNRS (Research Scientist) Information Genetique & Structurale UMR CNRS 1889 31, chemin Joseph Aiguier F-13402 Marseille Cedex 20

1 0