FEXH: find exons (human)

[Dan Davison]

      The Baylor College Of Medicine Computational Biology Group
			     Houston, TX
		       announces a new service

				 FEXH
			     (find exons)
				   
  Prediction of internal, 5'-  and 3'- exons in Human DNA sequences

NOTE: This service is temporarily being provided through the
University of Houston Gene-Server.  Only two jobs will be run at a
time.


Analysis of uncharacterized human sequences is available by sending a
file containing a sequence name as the first line and a sequence (with
no more than 80 characters/line) to

	        service at theory.bchs.uh.edu

with the subject line "FEXH". 

Example: mail -s FEXH service at theory.bchs.uh.edu < test.seq

where test.seq a file with the sequence.
 
Method description:
**********************
 
   Algorithm firstly predicts all internal exons in a given sequence   
   by linear discriminant function combining characteristics 
   describing donor and acceptor splice sites, 5'- and 3'-intron
   regions and also coding region for each  open reading frame  flanked 
   by GT and AG base pairs. Potential 5'- and 3'- exons are predicted
   by corresponding discriminant functions on the left side of the 
   first internal exon and on the right side from last internal exon,
   respectively.

Accuracy:
***************

  The accuracy of exon recognition have been estimated for a set 
  of 1016 exons from 181 complete genes.
  It contains nucleotide sequences from -150 bp before the first coding
  region and until +150 bp after the last coding region.
 

       		 Test:        Fexh        Grail-2   

  Exact exon prediction        70%        40%	    

  Exon nucleotides             85%(0.84)  77%(0.76)              

	The numbers in () are the correlation coefficients.

  It must be mentioned that this program does not assemble the predicted exons 
and it is more reliable for a case of exon missing (for example due to sequence 
errors). For a gene model prediction you can use "fgeneh" program from the 
Gene-Server (it has a better accuracy for complete gene structure prediction); 
or if you have only internal part of a gene sequence, internal exons may be 
predicted by server "hexon"  program.


Submitting sequences via email:
********************************
  For email submission the sequences must have the following format:  

Name of the sequence
ccatctctgtcttgcaggacaatgccgtcttctgtctcgtggggcatcctcctgctggca
ggcctgtgctgcctggtccctgtctccctggctgaggatccccagggagatgctgcccag
aagacagatacatcccaccatgatcaggatcacccaaccttcaacaagatcacccccaac
ctggctgagttcgccttcagcctataccgccagctggcacaccagtccaacagcaccaat
atcttcttctccccagtgagcatcg...............
   (Restrict the line length to 80 characters or less).


   You have to  send the file containing the sequence to: 

   service at theory.bchs.uh.edu
   Subject line must be:
   fexh
 Example: mail -s fexh service at theory.bchs.uh.edu < test.seq

Fexh output:		
*******************************
   1st line - name of your sequence
   2nd line - length of your sequence
   3d line - number of potential exons
   4th line and next - positions of predicted exons and their weights 

   For example:
   HUMALPHA     4556 bp ds-DNA             PRI       15-SEP-1 
   length of sequence -   4556
   number of potential exon:  10
   380 -    516   	w= 9.10
   611 -    727		w=11.10
   839 -    954 	w=12.33
  1147 -   1321		w= 7.70
  1819 -   1953 	w= 7.90
  2053 -   2125		w=12.51
  2254 -   2388		w =6.66
  2470 -   2661 	w=10.11
  2881 -   2997		w= 8.87
  3120 -   3562 	w= 9.92

Reference:

  1. Solovyev V.V.,Salamov A.A., Lawrence C.B.
   Predicting internal exons by oligonucleotide composition and 
   discriminant analysis of spliceable open reading frames. 
   (Nucl.Acids Res.,1994, in press).
  2. Solovyev V.V., Salamov A.A., Lawrence C.B.
   The prediction of human exons by oligonucleotide composition and 
   discriminant analysis of spliceable open reading frames.
   in: The Second International conference on Intelligent systems
   for Molecular Biology, (eds. Altman R., Brutlag D.,Karp R., Latrop R. 
   and Searls D.), AAAI Press, Menlo Park, CA 1994, (in press) 

Problems, comments, and suggestion:
   can be mailed to solovyev at cmb.bcm.tmc.edu.