Abundance and Variance In Taxonomy (‘AVeIT)

NGS has revolutionized explorations and analysis in the fields of microbiology, microbial ecology, contributing to our deeper understanding of health, and disease. The quality and quantity of processed sequencing data in the form of microbial taxa to represent diversity estimates of microbial communities, which is trending towards increasing number of samples and increase in sequencing depths, are in need of further attention and scrutiny. We present ‘AVeIT (Abundance and Variance In Taxonomy), an unbiased method to use all the available metrics (i.e. relative abundances in a sample, variance in a sample and abundance and variance in the whole study) in unison to identify members of microbial communities which could be potentially erroneous. In comparison to currently available strategies, at the level of taxa’s, in ‘AVeIT we systematically and reproducibly use the inherent abundance and variance of taxa in a given dataset, instead of using a priori thresholds, to determine the inclusion or rejection of taxa for further downstream analysis. While on one hand, our framework provides inferences that cannot be achieved using preexisting approaches; it does not affect the conclusions from traditional downstream analysis like PCoA.