16S V4 reads contain high % reverse primer

[Afina M]

Hi everyone,

I am running BI analysis for 16S V4 samples that was run using 515F/806R, MiSeq 300 PE. In average, all samples have more than 200k reads.

I run bduk below to remove the primers and more than 90% reads are trimmed out which only left 20k reads which I find it weird. I checked the stats output from bbduk and all samples contains really high % of reverse primer.

bbduk.sh in1=${SAMPLE}_R1_001.fastq.gz in2=${SAMPLE}_R2_001.fastq.gz \
		out1=ra_${SAMPLE}_R1.fastq out2=ra_${SAMPLE}_R2.fastq \
		ktrim=l k=$numk mink=$mink copyundefined=t \
		literal="GTGCCAGCMGCCGCGGTAA,GGACTACHVGGGTWTCTAAT" hdist=1 stats=${SAMPLE}_stats.txt \
		tpe tbo

Can anyone help to advice if this is normal? I run the tool multiple time with another 16S V4 dataset and the same thing happened. I did compare with 16S V3-V4 dataset and it only has around 40-50% primers.

I am still not really familiar with sequencing workflow so can anyone help me how to check this?

Thank you.