Non-specific amplification of human DNA is a major challenge for 16S rRNA gene sequence analysis

The targeted sequencing of the 16S rRNA gene is one of the most frequently employed techniques in the field of microbial ecology, with the bacterial communities of a wide variety of niches in the human body have been characterised in this way. This is performed by targeting one or more hypervariable (V) regions within the 16S rRNA gene in order to produce an amplicon suitable in size for next generation sequencing. To date, all technical research has focused on the ability of different V regions to accurately resolve the composition of bacterial communities. We present here an underreported artefact associated with 16S rRNA gene sequencing, namely the off-target amplification of human DNA. By analysing 16S rRNA gene sequencing data from a selection of human sites we highlighted samples susceptible to this off-target amplification when using the popular primer pair targeting the V3-V4 region of the gene. The most severely affected sample type identified (breast tumour samples) were then re-analysed using the V1-V2 primer set, showing considerable reduction in off target amplification. Our data indicate that human biopsy samples should preferably be amplified using primers targeting the V1-V2 region. It is shown here that these primers result in on average 80% less human genome aligning reads, allowing for more statistically significant analysis of the bacterial communities residing in these samples.