ABSTRACT
One of the most widely used techniques in microbiota research is 16S-rRNA-sequencing. Several laboratory processes have been shown to impact sequencing results, especially in low biomass samples. Low biomass samples are prone to off-target amplification, where instead of bacterial DNA, host DNA is erroneously amplified. Knowledge on the laboratory processes influencing off-target amplification and detection is however scarce. We here expand on previous findings by demonstrating that off-target amplification is not limited to invasive biopsy samples, but is also an issue in low bacterial biomass respiratory (mucosal) samples, especially when below 0.3 pg/µL. We show that off-target amplification can partly be mitigated by using gel-based library purification methods. Importantly, we report a higher off-target amplicon detection rate when using MiSeq reagent kit v3 compared to v2 (mean 13.3% vs 0.1% off-target reads/sample, respectively), possibly as a result of differences in reagents or sequencing recipes. However, since after bioinformatic removal of off-target reads, MiSeq reagent kit v3 still results in a twofold higher number of reads when compared to v2, v3 is still preferred over v2. Together, these results add to the growing knowledge base on off-target amplification and detection, allowing researchers to anticipate this problem in 16S-rRNA-based microbiome studies involving low biomass samples.
