DNA analysis programs

Bob Gross bob.gross at dartmouth.edu
Tue Mar 15 15:48:27 EST 1994

In article <9403140110.AA24617 at cscgpo.anu.edu.au>
Klaus.Matthaei at ANU.EDU.AU writes:

> Exactly my point it is supposed to compile overlapping sequences and
> assemble them as a contig and it couldn't for me, even when there was a
> 60bp TOTAL homology overlap.  I would be interested to hear if I am

What exactly do you mean by 60 bp TOTAL homology. If you mean that
there are 60 nts with 59 "gaps" in them my guess is that the program
would not (and should not) indicate an alignment. On the other hand, if
you have 60 bp in contiguous alignment the program should probably
indicate a match, but this still might not be part of a contig. Let me
try to draw this:

------------------------- DNA1
     /----------\         DNA2
    /            \

In this case a region of DNA2 matches a region of DNA1, but it would
have been impossible (without cloning artifacts) for these two
sequences to be part of a contig. The non-matched regions extend from
both sides of the match, so the original sequence fragments could not
have been overlapping in the genome (or wherever else you got the DNA).
This situation could arise if there are repeated sequences in your

It is not clear what the program should do in this situation. Any

Bob Gross
Dept. of Biology, Dartmouth College
bob.gross at dartmouth.edu

More information about the Methods mailing list