Interpreting Microbiome Analyses

High-throughput sequencing technology has made metagenomic-based studies possible. These studies use DNA sequencing to characterize the composition and potential function of a group of microorganisms from a specific environment. DNA from fecal samples is most commonly sequenced and classified against taxonomic databases to assess distinct compositional patterns. Three primary features of gut microbiota are evaluated¹:

Diversity

Microbial diversity is described as either alpha diversity (diversity within a unit) or beta diversity (differences in diversity between multiple units).

• Richness (number of species present), the Simpson index (measure of concentration of individuals of the same species), and the Shannon-Weaver index (measure of distribution of all species) are used to measure alpha diversity
• Beta diversity measures include the Bray-Curtis distance (differences in abundance of taxa between two communities), the Jaccard index (analogous to a qualitative Bray-Curtis distance that is naïve to relative abundance), and the UniFrac distance (measure of differences in the relatedness of two communities)

Composition

Amplicon and shotgun sequencing are used to characterize microbial composition. Most assessments characterize the connection between imbalances in distribution of commensal bacteria and disease rather than identifying specific relevant bacteria.

Function

At present, most studies do not regularly assess the actual function of gut microbiota. Assessing functionality requires multiple approaches and methods that can be difficult to perform. Advancements in technology are needed to help improve functionality assessments.

The tremendous volume of data produced by this research should be interpreted with caution. Most metagenomic studies assess fecal samples, and therefore the results predominantly represent luminal colonic microbial inhabitants and not microbiota from mucosal surfaces or other sections of the GI tract. Additionally, the results represent only a snapshot of a single point in time. These analyses also cannot differentiate between dead and live cells, which hinders the ability to distinguish between microbiota-disease associations vs causality.¹