A complex blast output sequence retrieval...

Przemko Tylzanowski przemko at sgi.celgen.kuleuven.ac.be
Tue Jan 9 04:00:53 EST 1996

Ther are a few programs that will allow for a more human analysis of
blast output than blast itself. Unfortunately none of them (to my
knowledge) would deal with the following problem.
I have a cDNA that, because is differentially spliced, has two ORFs. So,
to make my life easy (or so I thought) I searched databases using
tblastx. I got very nice local matches (blastp or blastn would not find
them) and decided to retrieve the sequences for more careful comparisons
(pileup, phylip etc.). Here comes the problem(s):
The blast output- unlike fasta- gives you only the matched part so if my
query has some gaps, I will get more than one alignement per sequence.
Since I have about 30 good hits, each of them give >1 local alignement,
That would not be so bad but tblastx dynamically translates database
sequences. Since most of my hits are against ESTs wich do not exist in
GenPep, altough I get the protein alignement on the screen, I cannot
retrieve the sequences per se (unless I copy and paste which is doable
but highly unpleasant).
So, the question is- is there a software that would allow me to take out
of the blast alignement the matches as independent sequences (no need to
go back to the databse and retrieve the original sequence). Then I could
use them as any other sequences in pileup, clustalw etc.
Przemko Tylzanowski Ph.D.

Celgen                                          phone:32-16-34-59-28
University of Leuven                               or:32-16-34-59-16
Herestraat 49                                     fax:32-16-34-59-33
3000 Leuven
e-mail:przemko at sgi.celgen.kuleuven.ac.besgi 5% 


More information about the Bio-soft mailing list