annotating EST sequence information with BLAST search output

Francois Jeanmougin pingouin at
Tue Feb 10 03:58:57 EST 1998

In article <199802100142.UAA06140 at>,
	francis at NCBI.NLM.NIH.GOV writes:
> true ... and this is probably why it's been so popular an activity.
> It's just that I'm not a great believer of automatic annotations ...

	Sure, because, sometimes, the second record contains more
information, and only a manual annotation would be useful in this

> yes, I agree ... it is faster, but what I question is what does it mean
> when you find you have something which is similar to somothing which is
> imilar to a homologue of something else?

	Yep, and what is the real information you are looking for in
an EST database? If you are looking for an homolog, you will find it
in the nr database. Looking for homologs of homologs, is, like Francis
says, quite a strange usage ;-).

	What is the real information you can find in a good EST 
database? For me :
	- Expression Pattern
	- Alternative splicing

	All the rest is quite unuseful. That's why I think that the good
way to maintain an EST database is by clustering homolog ESTs, like
STACKS do (but this implies better sequence quality or sequence cleaning).
	But probably, separatly, an expression database, a "size" or
cluster analysis one (to find full-length sequences or alternative
splicing), and whatever you would like. 

	But the current repository process is not as useful as it could be.

> **
> ** Be careful when reading DEFINITION lines of EST records, they can be
> ** informative, but they can also be misleading, and/or out of date!
> **

	Hey! stop increasing the size of the database! ;-)))). Try to
write it only with ATGCs, so we could compress it ;-).

	At least, end-user should finally stop using things 
without reading the docs...
					François (dreamin').
François Jeanmougin     | groupe de bioinformatique / bioinformatics groupe
tel:(+33) 3 88 65 32 71 | IGBMC BP 163 67404 Illkirch France

More information about the Bio-soft mailing list