Entering edit mode
6.0 years ago
ElUretsky
▴
30
I have 16s rRNA reads sequenced on an Illumina MiSeq 2x250bp. Our target region is about 300bp so pairs are expected to overlap. We used leeHom to merge paired-reads that overlap. Up to now it seems that we have good results but we are also getting many pairs where there is only a barcode combined with adapters, for example:
GCTCAGGA-AGATCGGAAGAACACACGTCTGAACTCCAGTCACGG....
TCCTGAGC-AGATCGGAAGAGCGTCGTGTAGGGAAAGAGTGTCTC...
The dash separates the adapter and the barcode. This is normally a minority of cases but there are some libraries where this is more frequent. Has else anyone experienced this? How can we prevent this for future runs?
Are you referring to primer dimers where there is no real insert? You would want to control that at the library prep/purification step. Bioinformatically you can scan and trim the data to remove the dimers (use
bbduk.shfrom BBMap suite, trimmomatic or cutadapt).they are very small so they do not make in the analysis pipeline. They question is more about how to avoid this maybe from groups doing 16s sequencing? sorry for the confusion.
You would generally do bead washing to remove those. This question is more appropriate for SeqAnswers.com if you want assistance with experimental strategy.