I am using a personal reference database to run the qiime2 feature-classifier classify-consensus-blast function on my sequences. It works fine. However, I was wondering if there was a way to get the percent identity log data. I am aware that if I choose a particular percent identity say(0.8) the query rejects the sequence that doesn’t meet that standard and they are marked unassigned. Currently the output .tsv file gives the confidence. What exactly does the confidence number means? Also, is there a away for me to get a log file output of percent identity of each of the accepted sequences(sequences over 0.8).
Not with classify-consensus-blast. q2-quality-control has an evaluate-seqs action that generates the log you are looking for, but it only reports the top hit (or rather the first reference sequence that exceeds the percent ID, as that is default blastn behavior). We do have an open issue to allow reporting more than one match with this action, but I do not have an eta for when we will get to that (but I can say it would be a really easy fix if you or someone you know has at least basic familiarity with python and would like to contribute )
Here it is the fraction of sequence hits (out of maxaccepts total hits) that agreed on the taxonomic lineage that is reported. It is the % consensus.
)