Remark about gap penalties

Guy Bottu gbottu at
Fri Jul 26 14:49:46 UTC 2002

from : BEN

	Dear colleagues,
I just noted something. Some EMBOSS programs (needle, water,...) use a gap 
penalty of type (<gap opening penalty> + (n - 1) * <gap extension penalty>),
while others (stretcher, matcher,...) use one of type (<gap penalty> + n *
<gap extension penalty>). The "align" libraries of EMBOSS, as far as I
understand, do not print the score computed by the algorithm but recompute
the score and always use the first type, what makes that the score they write
for e.g. matcher outputs is not correct. Shall this be fixed in a future
version or should programmers have the discipline of always using the first
type ?

Another remark : alignment programs write out a "help" text with line
  [-outfile]           align      (no help text) align value
Seems the standard message for "align" is missing.

	Guy Bottu

More information about the emboss-dev mailing list