Seqn alignment

brett at BORCIM.WUSTL.EDU brett at BORCIM.WUSTL.EDU
Thu Mar 17 11:33:55 EST 1994


Hello again. First, thanks to those who responded regarding interpretation
of sequence database search results. Apparently this is a topic that
needed clarification for us "wet" biologists, and I welcome any further
discussion on this thread.
 Now for a new question: What software should I use to produce an alignment
of many sequences. I am interested in aligning sequences from the genome
of my favorite virus in order to obtain a consensus sequence that would be
more useful as a query than any one sequence. As an RNA virus, it appears
to be highly variable  in certain parts of the genome. Lots of sequence
data exists in the banks (100's of entries), but most are fragments of the
genome, and a few are "complete" genomes. I work in a Mac-powered lab, but
have access a vax. In fact, I tried to begin this alignment project using
GCG, but have run into problems. Namely, PILEUP does not seem to work for
groups of sequences that contain only partial overlap, like this:

         +++++++++
    ++++++++++++++++
                         ++++++++++
 ++++++++++++++++++++++++++++++++++++
  +++++++++++++++++++++++++++++++++++

Which is how the sequence information exists. When I try this, PILEUP outputs
a blank file. However, PILEUP does seem to work when all the input sequences
are full-length genome. LINEUP seems good for aligning these internal overlaps
but can only handle 30 sequences. PRETTY, of course, requires a pre-existing
alignment as input. I have thought about aI thought about making several
alignments with LINEUP, and then lining up the consensuses, but I am afraid
that some residues might not be accurately represented. Will they? So, if
any of you have some ideas, I need to find a program that will:
     1 Mac or VMS based
     2 handle large numbers of sequences
     3 handle internal overlaps properly
     4 generate a consensus
     5 oh yeah, must be cheap too!

thanks in advance!
brett at borcim.wustl.edu




More information about the Bio-soft mailing list