IUBio

Homology Programs

mangalam at SALK-SC2.SDSC.EDU mangalam at SALK-SC2.SDSC.EDU
Tue Jun 2 13:39:01 EST 1992


Hi - I sent this once before, but it was in the thoes of setting up Eudora
and I may have sent it to my mother by mistake.  Sorry Mom...


 Luo Ning writes:

< lines deleted>

> 	I heard that there is a program called "Homology" run on
> Silicon Graphics.  I don't know if there are other similar
> programs available on net or on supercomputing centers such as
> Pittsburgh Supercomputing Center.
> 	I would appreciate very much if someone could enlighten
> me through E-mail.

<lines deleted>
  
   The program of which you speak is indeed called Homology and it does run
on on SGIs (4Ds, Crimsons, and Indigos with Elan Graphics) and IBM
RISCstations (I believe).  It is produced by BIOSYM (619 625 4555, 9685
Scranton Rd, San Diego, CA, 92121-2777).  

   It is definitely not free.  In fact, to run it you have to be running
the Insight/Discover package (also very much not free) of which Homology is
one of the many modules available.  That's the bad news.

  The good news is that it is a fabulous piece of software.  Much as the
initial piece described, it first does a modified FASTA search thru the
databases, using your sequence as a probe and then does an on-screen
alignment which you have the opportunity to modify.  I understand that they
are modifying the alignment interface (now a bit on the cantankerous side)
to be much more user-friendly.

  As long as the search turned up at least one member, the structure of
which has been solved, you can then do a 3-D alignment, using the solved
structure as a template.  Obviously, that will not be an accurate
representation because of the differences in side chains, spacing, etc, but
 you can then submit the initial model to Discover and run a series of
minimizations to more closely approximate the true conformation.  Or export
the structure and use XPLOR.

   One of the advantages of doing this is that even if the strucure is not
completely solved or accurate, you can certainly get a better understanding
of how a particular domain operates, especially after flipping it around
using Insight's phemomenal 3D graphics.

   Because of the cost of this package (hardware and software can easily
cost upwards of 100K), not everyone will be able to go out and buy it, but
most Supercomputer Centers do (San Diego does) and you can almost certainly
use it there.  One problem, though, is that Homology (and many of the
modules that make use of the graphics) requires the use of the console (ie
- although IRIX 4.x.x is X-windows based, you cannot export the window and
display it elsewhere (to my knowledge - I'd love to be proven wrong).  So
you or a proxy has to be sitting in front of a licensed workstation, at
least to do the initial set up.

   There was a note about it in a recent Nature (356:543 - SEM pic of
Coccolith Crystals on cover) by the head of the Homology development team
at BIOSYM, Rick Lee.  Interestingly, in the same issue just preceding the
above-noted piece was one by Desmet et al, reporting a dead-end elimination
algorithm for tremendously reducing the amount of computation required for
doing just these types of minimization.  I assume the BIOSYM people took
note.

[Std Disclaimers Apply]



Harry Mangalam                               Vox:(619) 453-4100, x250
Dept of Biocomputing                               Fax:(619) 552-1546
The Salk Institute                         mangalam at salk-sc2.sdsc.edu
10010 N Torrey Pines Rd                    mangalam at salk-sgi.sdsc.edu
La Jolla CA 92037                                mangalam at salk.bitnet





More information about the Bio-soft mailing list

Send comments to us at biosci-help [At] net.bio.net