In article <Pine.A32.3.91c.950123112943.26162A-100000 at homer22.u.washington.edu>, Peter Myler <mylerpj at u.washington.edu> writes:
|> On 22 Jan 1995, Mary Berbee wrote:
|>|> > We are looking for a program to tell us if two sequences are likely to be
|> > similar by chance or by homology. I assume there would be a method by
|> > randomly shuffling the bases (ie, keeping the base ratio the same but
|> > randomizing their order, then comparing the percent remaining the same
|> > when aligned).
|> > Any suggestions on this will be greatly appreciated!
|> > Dave Carmean carmean at sfu.ca|> > Mary Berbee berbee at unixg.ubc.ca|> >
|> There is a DOS program called RDF2 which does pretty much what you have
|> described. It used a FASTA type alignment, I believe. I don't know
|> where to get it on the NET, but I have a copy which I can send if you
|> want to e-mail me.
|> Peter J. Myler phone: (206) 284-8846x332
|> Seattle Biomedical Research Institute FAX: (206) 284-0313
|> 4 Nickerson Street e-mail: MYLERPJ at U.WASHINGTON.EDU|> Seattle, WA 98109-1651
If you do use RDF2 or any shuffling method to compare protein sequences
bear in mind that the mean SD score expected for UNRELATED proteins
is around 3.0.