Problem with classify.seqs

mothur bugs
1

I was running the classify seqs. command

classify.seqs(fasta=spore.trim.unique.good.filter.unique.precluster.pick.fasta, template=trainset6_032010.rdp.fasta, taxonomy=trainset6_032010.rdp.tax, cutoff=80, processors=2)

and at the end of the Mothur run I received a warning that some of the sequences may be reversed. However, I did use the flip parameter in the alignment to try to correct this problem. We looked in the accnos file that indicates which sequences are reversed and it appears that there are not that many of them so we were wondering if we could simply remove these sequences or continue on and Mothur would just disregard these sequences that are in reverse.

2

Mothur gives this warning when it is more probable that the reverse compliment’s kmers are in the template than the forward kmers. In that case the reverse compliment’s kmers are used for the classification. Since only a few of your sequences are falling into this category, I suspect they are just poor quality and may be removed when you run remove.lineage. You can also use remove.seqs to remove them if you want. Or you can select them with get.seqs, reverse them with reverse.seqs and add then add them back in with merge.files.

3

Thank you!