vsearch
vsearch copied to clipboard
Option to not ignore terminal gaps when computing consensus
A user has asked for this option:
When using "-cluster_fast" VSEARCH builds consensus sequences considering terminal gaps in the alignment, so a lot of relevant sequence information on the flanking regions is lost:
Example:
-----CCCAGT
----ACCCAGT
----ACCCAGT
ACTGAACCAGT
____________
ACCCAGT consensus
Could you add an option that builds the majority consensus but ignores positions wich have no coverage? so that the consensus sequence is:
ACTGACCCAGT
related to #25
Right now vsearch behaves like usearch and truncates the consensus when there is a majority of gaps, including terminal gaps. That's the default behavior. Should we introduce a --cons_notruncate
option to ignore terminal gaps when creating the consensus sequence? (see example above)