The adoption of sequencing-based technologies to probe biological and ecological questions has become exceedingly common due to significant reductions in cost, improvements in the quality of the data produced, and the ability to efficiently and quickly characterize many samples at a time, as well as the concurrent development of numerous bioinformatics tools, software packages, and reference databases. While great advances have been made, there are a number of shortcomings that remain as barriers to microbiome analysis being a precise, replicable, and reproducible science.<br/><br/>One of these shortcomings results from how raw sequencing data are processed to create counts of observed microbial abundances. Typical microbiome analysis pipelines approach read-to-taxon assignment as a discrete classification problem, wherein each read is explicitly assigned to a single member of an operational taxonomic unit (OTU) set. These approaches treat the data as though there is zero uncertainty in the assignments once they are made and, thus, they overestimate the certainty we have in any given read-to-OTU assignment. This leads to the second major shortcoming: a lack of appropriate computational and statistical methods to interpret the OTU tables. In some cases, difficulties and imprecision arise from using analytical approaches which were not developed with the nuances of sequencing data in mind, such as the data being non-census and compositional in nature.<br/><br/>In this work, we illustrate a method for incorporating uncertainty into typical microbiome analyses in order to overcome these challenges via a new and fully integrated bioinformatics and statistical pipeline. We apply this pipeline to two disparate datasets (fire-affected soil and photobioreactor) and demonstrate results obtained from a collection of statistics and hypothesis tests commonly used in microbiome research, including several alpha- and beta- diversity measures, differential abundance testing, and estimates of relative abundance for use in community profiling. Quantifying and accounting for uncertainties in the generated results reduces the risk of false positives due to understated variability, and improves the sensitivity of the tests, while making microbial community analysis more replicable in the process.

Propagation of uncertainty improves the quality of microbiome data analysis

Macromolecular Chemistry: The Second Century

[Author Benefits](https://scimeetings.acs.org/?utm_source=pubs_content_marketing&utm_medium=website&utm_campaign=0320_MCF_NPI_Launch_Spring_Homepage&ref=pubs_content_marketing)

[How to use SciMeetings](https://storage.googleapis.com/dl.morressier.com/acs/SciMeetings-ACSSpring2020-HowToGuide_M3020-Final.pdf)

[Frequently Asked Questions](https://storage.googleapis.com/dl.morressier.com/acs/SciMtng_ACSNMSp2020_FAQ_M3120_Final.pdf)

ACS Spring 2020 National Meeting & Expo

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.