Andrew Blakney

Comparing methods to assess the biodiversity of soil bacteria

Andrew Blakney, Simon Morvan, Marc St-Arnaud, Mohamed Hijri

Plant productivity and community composition varies according to the microbial biodiversity in the soil. This diversity can be monitored using Next-Generation Sequencing: DNA is extracted from the soil of a plant’s root system, and genetic barcodes for the microbes of interest, such as 16S for bacteria, are amplified by PCR. These amplicons can then be sequenced, and the reads analysed via a bioinformatics pipeline. To mitigate any errors that may have been introduced to the sequence data by PCR, or the sequencing reaction, retained sequences have typically been grouped together into operational taxonomic units (OTUs) based on 97% to 100% similarity. OTUs are usually then matched, or clustered, together into taxonomic groups, capturing the microbial biodiversity represented in the sequence data. However, OTUs simultaneously have high false positive rates—they overestimate diversity—and high false negative rates, by being unable to accurately discriminate real biological diversity from errors near the OTU-defining cut-off. Moreover, OTUs identified in different studies are not directly comparable. The recent DADA2 pipeline, however, identifies amplicon sequence variants (ASVs) with as little biological variation as one or two nucleotides, and correctly discriminates real diversity from errors. Since sequences can always be compared between analyses ASVs may be more valuable for future biodiversity studies. Thus, an objective of my PhD project is to evaluate how previously generated OTU data from plant roots relates to the ASVs identified in the same data by DADA2.