- modeller_usage - salilab.org

selenium parametrization
by Fabrizio Giordanetto 27 Oct '00

27 Oct '00

Hi, I got selenomethionines on a template structure and the program does not find selenium in topology library. Have you got selenium parametrized (radii...) or any suggestions in my next moves? thanks a lot. fab

1 0

SEQUENCE_SEARCH
by Troy Wymore 27 Sep '00

27 Sep '00

Hello MODELLERs, Could someone inform me on how MODELLER performs the search for templates given a sequence? I could not tell from the manual what algorithm it was using to do this. Thanks, Troy Wymore

2 1

maxbat
by Oliver Hucke 20 Sep '00

20 Sep '00

Dear modellers, I am trying to model a non-heme iron center within a protein. The iron molecule has 5 ligands (4 nitrogens from His-residues and a Glu sidechain) and 6 bonds to these ligands have to be considered. I succeeded with the reproduction of the binding geometry of a single his or the glu-sidechain by definition of a his and a glu patch. But trying to introduce all necessary 5 patches simultaneously, I get the following terminating error message in the log file: mkbnds__E> # of bonds/atom wrong; possibly increase maxbat atom indices in bond: 49 55 their bonds : 3 6 So, six bonds seem to be too many for Fe according to modeller ...? Can anybody help me with this. I did not find anything about "maxbat" in the modeller manual. How can I increase "maxbat" ? Thanks in advance. Best regards, Oliver. -- ******************************************* Oliver Hucke Institut fuer Physikalische Chemie II Universitaet Freiburg Albertstr. 23a 79104 Freiburg Germany Tel.: ++49-761-203-6179 Fax : ++49-761-203-6189 *******************************************

1 0

Re: Protein:DNA complexes in Modeller
by Andrej Sali 24 Aug '00

24 Aug '00

Mike Sierk wrote: > > Dr. Sali, > > We do not have Modeller installed here, and I was wondering how it does > modelling DNA-binding proteins. I have a structure of a protein:DNA > complex and would like to make a model of a related protein. Would it > work to just use the protein structure as a template, ignoring the DNA? Yes. > Could it model the unknown protein on the known DNA structure? Yes, but not with the default protocol. > Would it > model the unknown protein:DNA complex(with a different DNA sequence) > using the known complex? Yes, but not with the default protocol. The results are not likely to be overwhelming. Best, Andrej > > I found only two questions about DNA in the archive, with no answers to > either. > > Thank you, > Michael Sierk > > -- > > ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ > Michael Sierk email: mls5w(a)virginia.edu > Pharmacology Dept./Biophysics Program Phone(W): (804) 243-6725/924-1953 > Box 800735, Health Sciences Center (H): (804) 970-2268 > University of Virginia Fax: (804) 924-9057 > http://www.people.virginia.edu/~mls5w > GBS Homepage: http://www.med.virginia.edu/ed-programs/gpo/gbs/ > Grad Council Homepage: http://www.student.virginia.edu/~gsasc/ > ------------------------------------------------------------------------ > "If at first you do succeed, try not to act surprised." > - Anonymous. > ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ -- Andrej Sali, Associate 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)rockefeller.edu; http://salilab.org

1 0

IRIX 6.5 problems; install patch 3371
by Andrej Sali 21 Aug '00

21 Aug '00

Problems with Modeller under Irix6.5: Patch 3371 is absolutely necessary. -- Andrej Sali, Associate 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)rockefeller.edu; http://salilab.org

2 2

residue type undefined
by J.U. Flanagan Biomedical Research Centre 18 Jul '00

18 Jul '00

Hi all, Could anyone tell me what this error message is indicative of, and how I can fix it? iup2crm_E> IUPAC atom not found in topology library: CG iup2crm_E> residue type undefined recover__> MODELLER_STATUS >= STOP_ON_ERROR: 1 1 Thanks in advance Jack

2 1

Re: Modeler - adding more models
by Andras Fiser 07 Jul '00

07 Jul '00

Hi Randa, OK, unless you set a different random start you will always have the same set of models i.e.. include in your top file e.g.: SET RAND_SEED = -10973 see, default: RAND_SEED = <integer:1> -8123 random seed from -50000 to -2 Andras RATHNACHALAM_RADHAKRISHNAN(a)Lilly.com wrote: > > Dear Andras, > > May be I was not clear when I asked my question. Initially I have made 5 > models. I wanted to make more models. > So I made the changes that I had indicated in my earlier message. Now I do > get models from 6 to 10. They are > named correctly. Only thing is the model 6 was the same as model 1, model > 7 was the same as model 2 and like wise. > How can I save time without redoing everything again? Thanks for your > reply. > > radha > > Andras Fiser <andras(a)tuba.rockefeller.edu>@mail.rockefeller.edu on > 07/07/2000 02:06:28 PM > > Please respond to fisera(a)rockefeller.edu > > Sent by: andras(a)mail.rockefeller.edu > > To: RATHNACHALAM_RADHAKRISHNAN(a)lilly.com > cc: modeller_usage(a)salilab.org > > Subject: Re: Modeler - adding more models > > It should work as you suggested, I suspect some trivial syntax problem. > > Andras > > RATHNACHALAM_RADHAKRISHNAN(a)lilly.com wrote: > > > > Hello: > > > > If I have already made say 5 models using Modeler, and if I would like > > to make models say from 6 to 10 what should I do. I tried to edit the > top > > file > > and changes the STARTING_MODEL and ENDING_MODEL TO 6 and 10. > > It seems the output files are the same. Is there another control to let > > the program > > take a different path to get more models? Thanks > > > > radhakrishnan -- , Andras Fiser, PhD # phone: (212) 327 7206 The Rockefeller University # fax: (212) 327 7540 Box 270, 1230 York Avenue # e-mail:fisera@rockefeller.edu New York, NY 10021-6399, USA # http://salilab.org/~andras

1 0

Modeler - adding more models
by RATHNACHALAM_RADHAKRISHNAN＠LILLY.COM 07 Jul '00

07 Jul '00

Hello: If I have already made say 5 models using Modeler, and if I would like to make models say from 6 to 10 what should I do. I tried to edit the top file and changes the STARTING_MODEL and ENDING_MODEL TO 6 and 10. It seems the output files are the same. Is there another control to let the program take a different path to get more models? Thanks radhakrishnan

2 1

Re: Modeller - Chain IDs
by Andras Fiser 02 Jul '00

02 Jul '00

Hi Thorsten, Once you are done with modeling, make another top file like this one: ######### READ_MODEL file='LONGSEQ.pdb' RENAME_SEGMENTS SEGMENT_IDS = 'A' 'B' 'C' , RENUMBER_RESIDUES = 1 3 3 WRITE_MODEL FILE='SEGMENTED.pdb' ########### This will take the resulting model with 1000+ residues (LONGSEQ.pdb: continuously numbered and no chain IDs), but TER records included as you defined in the alignment file by '/'. Next, in the above top file I assumed that there are 3 segments, they will be labeled as A, B, C and each segment will be renumbered and start with the specified starting numbers, in this case with 1, 3 and 3 respectively. So, look at the RENAME_SEGMENT command for details. Andras If you do not mind I am sending this reply to the list as well, since this question was not discussed so far. > Dear Andras, > > I have a question concerning the modelling of my hexamer > with the modeller version you gave me: > > I used the hint on the mailinglist concerning modelling > hetero-trimers in 1998 - to use the slash / to identify a > new chain - it works fine, but Modeller does not insert > any chain IDs... (there are IDs in the template). > This means the aa-id-numbers increase to over 1000, and > in the PDB-format, only three numbers are allowed for one > chain, so the visualization programs cut off the first number > and start again with 000... > > How do I get the chain-IDs and a correct aa-numbering? > > Thanks, > Thorsten -- , Andras Fiser, PhD # phone: (212) 327 7206 The Rockefeller University # fax: (212) 327 7540 Box 270, 1230 York Avenue # e-mail:fisera@rockefeller.edu New York, NY 10021-6399, USA # http://salilab.org/~andras

1 0

a problem
by Matteo Dal Peraro 16 Jun '00

16 Jun '00

in a modelling run I have this error during: 109 MAKE_RESTRAINTS RESTRAINT_TYPE = 'chi1_dihedral' ADD_RES; TRAINTS = ON Restraint type to be calculated: chi1_dihedral getpdf__E> your library requires all bins zero; nbinx: 4 what does it mean, I didn't find informations about it how can I resolve it? thanks matteo

2 1