[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[ccp4bb]: Homology modelling summary




The QUESTION:   I was wondering what is the easiest way to apply a homologous sequence to a known crystal structure to create a homology model?

What fun it was to post a question on ccp4, here are the results.  Not necessarily in the order they came in.  MY COMMENTS AND STUPID JABS ARE IN CAPITOLS.

Gerard DVD Kleywegt <gerard@xray.bmc.uu.se>  i tend to use Modeller for this, see
http://guitar.rockefeller.edu/modeller/modeller.html but there are also various web-based servers to which you only need to submit a sequence, see e.g. ProSAL for a quick way to do this http://xray.bmc.uu.se/sbnet/prosal.html)

kline <K.Line@exeter.ac.uk> A very easy way, if the homology is good is just to submit the sequence to  SwissProt modeller over the web, and it will send you the structure and the
stats!"  Eric J. Haas" <ehaas@molbio.unmc.edu> I've worked with SwissModel via SwissPDB Viewer many times. Hurley, Thomas D." <thurley@iupui.edu> I would use the program MODELLER from A. Sali for this purpose, alternatively, given a good alignment SPDV (Swiss-PDB Viewer available from Swiss-Prot) has a good tool for this as well. <phubbard@post.its.mcw.edu> MODELLER, THREADER2, SPDB (SWISSMODELLER)  The list goes on... I have used MODELLER a couple of times, and it seems to work pretty well (comparing to the subsequently solved structure), though it is pretty picky about data formats.  Gregory Verdon <g.verdon@chem.rug.nl> There is SWISS-Model:  
http://www.expasy.org/swissmod/SWISS-MODEL.html  Phoebe Rice <price@midway.uchicago.edu> Try swissPDBviewer - you can see & adjust th! e alignment in one window and watch the new side chains sprout off the old Ca's in a 2nd window. Nic Steussy <csteussy@purdue.edu> I've had good luck with the SwissPDB viewer.  The Tutorial has excellent instructions.  http://us.expasy.org/spdbv/

zeth <zeth@biochem.mpg.de> http://www.cmbi.kun.nl:1100/WIWWWI/
 
Michael Chan <chan@chemistry.ohio-state.edu> Try SOD within the O suite of programs.  

BUT I'M ALLERGIC TO O!!

Wen Hwa Lee <leewh@scripps.edu> A little less known program is ICM, which makes a great job as well. It is a paid package, but the author (Dr. Ruben Abagyan, from Scripps) made the automated homology modelling module freely available as a web server:  http://abagyan.scripps.edu
(look for item #5.3 homol. model. server on a frame window on your left)  It automatically looks for the best template and builds the model, that can then be retrieved as a PDB file.


Xiaodong Zhang <xiaodong.zhang@ic.ac.uk>  Another excellent web-based program is 3D-JIGSAW.
http://www.bmm.icnet.uk/servers/3djigsaw/

THEN THE DAY BECOME MORE INTERESTING

This in from Bart Hazes Bart Hazes <bhazes@ualberta.ca>  I just happened to come accross the following reference:  Increasing the precision of comparative models with YASARA NOVA--a
self-parameterizing force field.  Krieger E, Koraimann G, Vriend G. (Proteins 2002 May 15;47(3):393-402)  One of the conclusions drawn at the CASP4 meeting in Asilomar was that applying various force fields during refinement of template-based models tends to move predictions in the wrong direction, away from the experimentalLy determined coordinates. etc... "I have been recommending people not to make homology models at all. Just use the template as is and your brain to extrapolate from the template. that I should not bother with a homology model"  

BUT ... MY BRAIN ISN'T BIG ENOUGH.

BARTS COMMENT STARTED AN INTERESTING DISCUSSION.

Randy Read <rjr27@cam.ac.uk>.... My (also biased?) impression is that homology modeling makes the models worse for molecular replacement, and I have been suggesting that it's probably best
to stick to the original model, omitting only parts that are clearly very different, and perhaps downweighting loops and side-chains that are likely to fit poorly (as done in MolRep). ... It would be interesting if people could supply clear-cut examples of where homology modeDoes anybody out there have any such examples?  Is my impression still correct, or is it finally out of date?

OR HAS ALTZHEIMER'S FINALLY SET IN??? (just kidding Randy)

Karsten Suhre <Karsten.Suhre@igs.cnrs-mrs.fr> Karsten: In reply to Randy Read's question: Yes, we have recently solved a structure using homology modelling in a situation where molecular replacement failed (i.e. the two potential templates had less than 25% sequence identity). A second case also would have worked, but as we had MAD data available for that protein at that time we did not proceed any further on it. However, it did take some time to get there, that is testing different alignments (T-COFFEE-SAP worked best) and scoring numerous different models (using MODELLER) with AMoRe before we finally got one that worked (there were 4 molecules in the asymmetric unit, each 200 residues long). We also erased a badly aligned loop and some of the N-terminal residues, but they could be reconstructed using ArpWarp and Buster lateron. Note that D.T. Jones proposed a similar approach in Acta Cryst. (2001) D57, 1428-1434.

Gerard DVD Kleywegt <gerard@xray.bmc.uu.se> since the proof of the pudding is in the eating,
OH GREAT NOW I'M  HUNGRY
we ate some while evaluating the comparative modelling bit in CASP3 (autumn 1998), using a target we had submitted ourselves and for which we therefore had the experimental data.  at that stage essentially all homology models performed worse than the template (either with its own, incorrect sequence, or turned into a poly-Ala). see: T A Jones & G J Kleywegt, Proteins, Suppl. 3, 30-46 (1999) [get a reprint via http://xray.bmc.uu.se/cgi-bin/gerard/reprint_mailer.pl?pref=53]. having said that, once you're faced with a near impossible molecular replacement problem (and believe me, i've been there), you're probably going to want to grab every straw, and that could well include homology models. after all, you haven't got much to loose (apart from some time), and everything to gain.

HOPE for the future:
Lesa J. Beamer" <lesa@bluebird.biochem.missouri.edu> Just today I ran across the following web site:  http://guitar.rockefeller.edu/modbase
where Andrej Sali and coworkers have compiled more than 500,000 homology models for sequences in Genbank, based on PSI-blast searches against the available crystal structures in the PDB.   Haven't had time to look carefully at any of them yet, but maybe none of us will have to build homology models again.  


Pat Loll <pat.loll@drexel.edu>, was very helpful, don't be lazy, just do the crystal structure!

TO BE HONEST, I WASN'T GOING TO DO MOLECULAR REPLACEMENT AT ALL!  I JUST WANT TO HELP A COLLEAGUE UNDERSTAND WHAT A RECENT STRUCTURE FROM ANOTHER SPECIES MEANT FOR THEIR RESEARCH ON THE HUMAN COUNTERPART.  

I HAVE TO GO PLAY WITH swissPDBviewer NOW, AND THEN SOME OF THE OTHERS.

THANK YOU FOR THE HELP.
******************************************************************************
Gloria Borgstahl                                            
Eppley Institute for Cancer Research
987696 Nebraska Medical Center  
10732A Lied Transplant Center              
Omaha, NE 68198-7696

Office (402) 559-8578
FAX (402) 559-8577

Associate Professor                
Interests:  Cancer, Biochemistry, DNA Metabolism,
Protein Crystallography, Super-fine phi slicing, Synchrotron
Radiation, Crystal Mosaicity, and X-ray Topography
******************************************************************************