[Biopython-dev] [Bug 1876] Bio.pairwise2 generates incorrect Needleman-Wunsch score_matrix

[bugzilla-daemon at portal.open-bio.org]

http://bugzilla.open-bio.org/show_bug.cgi?id=1876





------- Comment #8 from bill at barnard-engineering.com  2005-10-12 13:40 -------
FWIW your traceback algorithm works perfectly as is for the Gotoh/Durbin/NW
algorithm. The internal trace_matrix is somewhat different since each pointer
may only point to three possible cells rather than to the previous max score
cell (IIRC...).

Anyway I used your algorithm in my program. (I want to be sure I understand all
the basics before I move to something more challenging.) I tried both
extracting the portions I needed, and simply importing it directly; both work
perfectly for my example alignment.

I fooled around a bit with making test routines to compare the output of the
Gotoh algorithm to the NW algorithm. I learned a bit, but I don't think that
adding these tests to the test_pairwise2 module would really add anything
useful.

Thanks for making all this publicly available. I really like the way your
program uses the __call__ ==> decode methods to enable flexible use of the
alignment programs. It's opened my eyes to Python's capabilities.




------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.