299. Microbiome Analysis as a Tool to Investigate Gut Health: Methodological Considerations.

Abstract

Molecular tools are used to assess intestinal microbiota, with next-generation sequencing (16S rRNA gene or DNA shotgun sequencing) being the most applied methods in research. These untargeted methods are particularly useful for discovering new bacterial taxa and identifying broad microbial changes associated with disease or nutritional interventions. However, they have significant limitations: major challenges include the lack of standardization and the high variability between laboratories, as differences in DNA extraction, choice of PCR primers, statistical models, and bioinformatics pipelines (ie, databases) all influence the results. Consequently, even when analyzing the same dataset, different laboratories report different findings. This contributes to the variability observed across studies, even when similar diets or health conditions are examined. A major limitation is that sequencing data are reported as relative abundances rather than absolute quantities. Because each taxon is expressed as a proportion of the total sequences, an increase in one bacterial taxon will automatically appear as a decrease in others, even if their actual levels remain unchanged. For example, if Streptococcus increases in response to a dietary change, other taxa may be reduced simply due to this proportional shift, not because they truly declined. As a result, reporting data as relative percentages leads to inaccurate quantification of microbiome changes. Additionally, sequencing methods generally have lower analytical sensitivity, meaning some important bacteria remain undetected. In contrast, targeted quantitative PCR (qPCR) offers higher sensitivity and reproducibility, allowing for more accurate detection of specific taxa. Each bacterial taxon is measured independently, therefore avoiding the pitfalls of compositional data (relative abundance). This allows us to establish reference intervals, and furthermore, because assays are reproducible, to compare data across studies. This presentation will discuss the common methodological pitfalls of microbiome studies, and how evaluation of core bacteria and their functional metabolites using standardized and targeted assays improve the interpretation of microbiome data.