Everyone! When I dealt with a 16S single end dataset produced by 454 FLX, I found that most reads have primer and barcode, and some reads just have barcode and no primer. With q2-cutadapt, I trimmed the primer and the proceeding bases and at the sametime I discard the untrimmed sequences. I found about 5% of all reads were discarded. Is it ok? Will this affect the downstream analysis and authenticity?
That depends — is this what you expect? This all sounds very fishy to me and unless if this is expected based on your library prep protocol I recommend contacting whomever provided these data to determine why many reads are missing the primer/barcodes. So far this does not really sounds like a QIIME 2 issue per se that we can assist with.
@Nicholas_Bokulich Thank you for your answer！