[mira_talk] Re: Smith-Waterman algorithm
- From: Laurent MANCHON <lmanchon@xxxxxxxxxxxxxx>
- To: mira_talk@xxxxxxxxxxxxx
- Date: Sat, 22 Nov 2014 11:03:26 +0100
Le 22/11/2014 08:43, Bastien Chevreux a écrit :
On 21 Nov 2014, at 15:13 , Laurent MANCHON <lmanchon@xxxxxxxxxxxxxx
<mailto:lmanchon@xxxxxxxxxxxxxx>> wrote:
is MIRA use the fast implementation of the Smith-Waterman algorithm
(https://github.com/mengyao/Complete-Striped-Smith-Waterman-Library)
during the calculation of alignments ?
I actually played around with that SW implementation. The SW search
itself is blazing fast (for a SW search), the big drawback there is
the backtracking algorithm which produces the aligned strings. Taking
these two steps together, the implementation above and the MIRA
version are more or less equivalent in speed so I scrapped plans to
adapt it for MIRA.
Note that I tested this one or two years ago, maybe the authors made
improvements I don’t know of.
B.
...it was just a suggestion to improve alignment step that consume time.
Laurent --
Other related posts:
