[BiO BB] Re: a validation study

l x yi lxyiwc at yahoo.com
Tue Dec 7 10:45:02 EST 2004


I was reading some references, but different people
were using different datasets. I got confused. For
example, to simulate random sequences, there are at
least several ways: 
-- simulate sequences with frequences of each of 20 aa
as in SWISS-PROT
-- simulate seq freq according to Robinson and
Robinson (1991), PNAS, 88, 8880-4 by BLAST  paper.
-- simulate seq freq by McCaldon et al. (1988)
oligopeptide biases in protein seq and their use in
predicting protein coding regions in nucelotide
sequences. 

also, for the set of profiles, one way is to use the
top 20 seed alignment of profiles in
pfam,http://pfam.wustl.edu/browse.shtml
but there are always several sections of the profiles,
could I randomly cut out a section of a profile from
each of the top 20 profiles? see 
http://pfam.wustl.edu/cgi-bin/getalignment for
example. 

Thanks so much for all the suggestions. 

Lily







		
__________________________________ 
Do you Yahoo!? 
Read only the mail you want - Yahoo! Mail SpamGuard. 
http://promotions.yahoo.com/new_mail 



More information about the BBB mailing list