datbase

[Keith Vass]

MIcha Ron wrote:
> 
> Dear netters,
> 
> I have a question about GCG software.  How do I create a new subset of
> sequences from an output of stringsearch.  I would like to do a fasta
> search with all the output sequences recieved from the stringsearch.
> 
> Mark Band
> michar at indycc1.agri.huji.ac.il
Francois Jeanmougin wrote:        Like lookup, stringsearch gives you a
file of sequence names (FOSN)
formated output. So you can use fasta as follow :
--
fasta my_sequence
(...)
Search for query in what sequence(s) (* SwissProt:* *) ?
@my_subset.strings
---
the @ says to GCG that this is a FOSN file, and GCG will take all the 
sequences to make the search . You don't need to create a database,
nor to copy all the sequences (no need of disk space).

This is completely Sensible. 
But it is useful to define a SUBSET of the sequences identified by
stringsearch. In unix, to identify all mouse sequences in genbank names
you can do this with a file called mydata.strings:
 
egrep -i mouse mydata.strings > mymousedata.strings

this will put all the mouse entries into a new file called 
mymousedata.strings.

Keith Vass