Blast search help needed.

William R. Pearson wrp at alpha0.bioch.virginia.edu
Thu Oct 3 11:26:23 EST 1996


The current versions of the FASTA package, which includes FASTX, will
allow you to search subsets of a DNA query sequence sequence against a
database. You can specify the part of the query sequence you which to
use as shown below:

	fastx query.dna:1-99 swissprot

It is relatively easy to run a unix shell script that will search each
XXXX nt piece of a query sequence.  Here is one that I have used
(csh):

    #!/bin/csh
    set m=0
    while ( $m < 40000 )               # query sequence is 40000 nt long
    @ n = ($m + 1100)                  # search in 1100 nt chunks
    @ start = ($m + 1)
    if ( $start > 1000 ) then          # use a 100 nt overlap
    @ start = ($start - 100)
    endif
    echo "searching $start - $n"
    fastx3_t -q cosmid.seq:${start}-${n} s > cos.${start}-${n}.ok2x_s
    fastx3_t -i -q cosmid.seq:${start}-${n} s > cos.${start}-${n}.ok2y_s
    @ m = ($m + 1000)                   # move down 1000 nt
    end





More information about the Bio-soft mailing list