classify.seqs knn, low quality hits

Commands in mothur
1

Sorry if this is answered elsewhere, didn’t find it.

I am using the knn/blast method of classify.seqs and sometimes the “best” hits are bad hits (short, low %id) and result in incorrect assignments. Is there any way to set minimum parameters for a “best” hit?

Thanks,

Sean

2

Sorry, Sean there isn’t. How about using the Wang, naive Bayesian approach instead?

Pat