high percentage of unclassified sequences on Illumina

[Volunteer42]

Hi all,
I'm sequencing V6 region of bacterial 16S on an Illumina MiSeq and when I got my sequences back and process them using mothur, I'm getting a LOT of unidentified sequences. Like maybe 50% for some samples in the run. This wasn't the case for other Illumina runs on similar samples. Any ideas on what the cause may be? Is it appropriate to just remove these unidentified sequences from further analysis and publication or not? Your help is greatly appreciated.

[thomasblomquist]

QIIME pipeline a subset of your reads. A few "OTU" consensus sequences will come from this. Blast then against the NCBI nucleotide database and you'll have a general idea if it's technical artifact, an odd variant of your target of interest, or something completely different. -Tom

[bilyl]

FastQC will also give you a quick and dirty check of any "overrepresented" sequences. Depending on what you're trying to do, you might be picking up a lot of eukaryotic contamination or Illumina adapter dimers.