Naming motifs silently breaks consensus filtering
rsharris opened this issue · 2 comments
When NCRF is run with <name>:<motif>, instead of just <motif>, the resulting alignment file doesn't contain the motif's nt sequence. ncrf_consensus_filter needs that sequence to figure out whether the consensus matches it or not. Moreover, it needs it to figure out how to partition the alignment back into units.
Currently, ncrf_consensus_filter doesn't recognize that aspect of an alignment file.
Until that is fixed, the workaround is to not name motifs.
This has been corrected in the active state of the repository, but is not yet part of a tagged release.
The solution requires the user to provide ncrf_consensus_filter the same <name>:<motif> options that were given to the NCRF alignment command.
The fix is included in tagged release v1.01.02