-----Original Message----- From: andras@rockvax.rockefeller.edu [mailto:andras@rockvax.rockefeller.edu]On Behalf Of Andras Fiser Sent: January 27, 2000 10:39 AM To: Andrej Sali; kevin.truong@utoronto.ca Subject: Re: [Fwd: Modeller mysteries]AF: handling HETATM entries in the middle of PDB file

Hi Kevin,

In this strange pdb file the hetatm entries inserted at several intermediate positions in the coordinate file. You tried to include them by including X-es in your alignment but it does not work like this. At first, if you do not read in HETATM entries, which is the default situation, (look at command HETATM_IO) than Modeller skips these entries so you do not need to include them in your alignment file either. Second, even if you include them you should not stick on the PDB file residue numbering (it is often chaotic): for some reason in your case the numbering skips by three at the HETATM enty although there is only one HETATM entry. You included 3 X-es accordingly, you need only one. Modeller renumbers the file by simply reading in order all the coordinate entries from the file irrespectively what numbering is associated with the entries in the PDB file. Third, if you do include HETATM entries you should indicate it by "." and not by "X". If you just replace your X-es to "." and instead of XXX you use a single "." your setup already works for me.

I suggest you send your questions to the Modeller mailing list in the future, you might have answers from others and others also can learn. Please look at: http://salilab.org/cgi-bin/lwgate/lwgate/MODELLER_USAGE/

if you do not consider this mail as confidential please forward it to the list: modeller_usage@salilab.org

best wishes,

Andras > > Subject: Modeller mysteries > Date: Wed, 26 Jan 2000 19:51:03 -0500 > From: "Kevin Truong" kevin.truong@utoronto.ca > To: sali@rockvax.rockefeller.edu > > Hi Prof. Sali: > > My name is Kevin. I'm from Ikura's lab at the Ontario Cancer Institute. > I'm having some problems using modeller that I hope you can help me with. > I'm trying to model the RFP sequence with the GFP structure. > > I download the pdb file of 1ema from the PDB protein bank. > I ran the Modeller and it was not able to produce the predicted pdb file. > I have attached the following files: > > pdb1ema.ent pdb file used for Modeller > cFP484.ali ali generated by Threader > cFP484.top top script for Modeller > cFP484.log result of the Modeller run. > > I have no clue why this does not work. Can you help me? > > Thanks! > Kevin

-- , 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