Protein Structure Prediction Software

Andrew Martin martin at bsm.bioc.ucl.ac.uk
Fri Mar 4 04:55:12 EST 1994


In article <2l4skp$bur at fermat.mayo.edu>, eberhard at Mayo.EDU (norm eberhardt) writes:
|> 	I am aware of Chou-Fasman types of programs which predict secondary
|> structure on the basis of amino acid sequence.  Are there similar programs
|> available that model tertiary (3-d) structure on the basis of amino acid sequence
|> only?
|> 
|> Thanks,
|> Norman L. Eberhardt  
|> 
Unfortunately this is a rather simplistic request! If only there were
such programs we could put all the crystallographers out of a job :-)
Secondary structure prediction rarely does better than 60-70% accurate.

However, many people are working on this area using a number of homology based
methods and a fairly new technique of sequence threading (trying to find the
best tertiary fold to accomodate a sequence), but it will be a good few
years before one will get reliable results.

Now if you already know pretty much what your structure will look like (i.e. you
already have the structure of one, or preferably more, closely related
structures), that's a different matter and we can do pretty well.

Andrew
-- 
****************************************************************************
Dr. Andrew C.R. Martin,         University College London & SciTech Software
INTERNET: martin at bsm.bioc.ucl.ac.uk     -OR-     amartin at scitec.adsp.sub.org
JANET:    martin at uk.ac.ucl.bioc.bsm
UUCP:     {uunet|rutgers}!cbmehq!cbmuk!scitec!amartin
****************************************************************************



More information about the Proteins mailing list