Comprehensive biodiversity analysis via ultra-deep patterned flow cell technology: a case study of eDNA metabarcoding seawater

Abstract

The characterization of biodiversity is a crucial element of ecological investigations as well as environmental assessment and monitoring activities. Increasingly, amplicon-based environmental DNA metabarcoding (alternatively, marker gene metagenomics) is used for such studies given its ability to provide biodiversity data from various groups of organisms simply from analysis of bulk environmental samples such as water, soil or sediments. The Illumina MiSeq is currently the most popular tool for carrying out this work, but we set out to determine whether typical studies were reading enough DNA to detect rare organisms (i.e., those that may be of greatest interest such as endangered or invasive species) present in the environment. We collected sea water samples along two transects in Conception Bay, Newfoundland and analyzed them on the MiSeq with a sequencing depth of 100,000 reads per sample (exceeding the 60,000 per sample that is typical of similar studies). We then analyzed these same samples on Illumina’s newest high-capacity platform, the NovaSeq, at a depth of 7 million reads per sample. Not surprisingly, the NovaSeq detected many more taxa than the MiSeq thanks to its much greater sequencing depth. However, contrary to our expectations this pattern was true even in depth-for-depth comparisons. In other words, the NovaSeq can detect more DNA sequence diversity within samples than the MiSeq, even at the exact same sequencing depth. Even when samples were reanalyzed on the MiSeq with a sequencing depth of 1 million reads each, the MiSeq’s ability to detect new sequences plateaued while the NovaSeq continued to detect new sequence variants. These results have important biological implications. The NovaSeq found 40% more metazoan families in this environment than the MiSeq, including some of interest such as marine mammals and bony fish so the real-world implications of these findings are significant. These results are most likely associated to the advances incorporated in the NovaSeq, especially a patterned flow cell, which prevents similar sequences that are neighbours on the flow cell (common in metabarcoding studies) from being erroneously merged into single spots by the sequencing instrument. This study sets the stage for incorporating eDNA metabarcoding in comprehensive analysis of oceanic samples in a wide range of ecological and environmental investigations.

Non‐specific amplification compromises environmental DNA metabarcoding with COI

“…Particularly important for regulatory applications, or where researchers wish to compare results over time or between studies, some degree of standardization is desirable (Hering et al, ). Our results – and those of previous studies using similar primer sets (Bakker et al, ; Jeunen et al, ; Lim et al, ; Macher et al, ; Singer et al, ; Stat et al, ; Yang et al, ) – show that environmental metabarcoding for restricted taxonomic groups using degenerate COI primers results in excessive volumes of ‘wasted’ sequencing effort. This co‐amplification of prokaryotic and non‐target eukaryotic DNAs and subsequent lack of specificity is due to the nature of mutation patterns in COI (Siddall et al, ).…”

“…That the highly degenerate Leray‐XT primers produced a low proportion of fish reads is unsurprising given that previous studies on environmental samples using degenerate COI primers have demonstrated that they can amplify widely beyond their target taxa, and can produce large proportions of unassigned reads (Lim et al., ; Macher et al, ; Singer et al, ; Stat et al, ). The proportion of bacterial reads are generally lower when metabarcoding bulk organismal samples, however, with most reads belonging to metazoans (Leray & Knowlton, ; Macher et al, ; Wangensteen et al, ).…”

The utility of a metagenomics approach for marine biomonitoring

“…It is interesting to note that on Illumina DNA sequencing platforms-the most popular instruments for metabarcoding experiments at the moment (Singer et al 2019)-adding a metagenomics library to the flow cell can potentially be performed without adding any additional sequencing costs. This is because Illumina recommends spiking in a control library composed of PhiX genome to increase base complexity on the flow cell, with a recommended proportion of PhiX ranging from 5% to 50% (Illumina, 2017).…”

Shotgun metagenomics and metabolomics reveal glyphosate alters the gut microbiome of Sprague-Dawley rats by inhibiting the shikimate pathway

“…In addition, different classifiers have been shown to provide different results and there is no gold standard method by which to analyse shotgun metagenomics datasets (Ye et al 2019). Even the type of instrumentation employed can play a role, with NovaSeq sequencers detecting more DNA sequence diversity within samples than MiSeq sequencers at the exact same sequencing depth (Singer et al 2019). This could be amplified in our study by the housing of three rats per cage, since rats can exchange faecal microorganisms and metabolites by coprophagy (Suckow et al 2005).…”