Power of Microbiome Beta-Diversity Analyses Based on Standard Reference Samples

• Activity Number: 251 - Statistical Advances in Microbiome Research from Theory to Application
• Type: Invited
• Date/Time: Tuesday, August 4, 2020 : 1:00 PM to 2:50 PM
• Sponsor: Section on Statistics in Genomics and Genetics
• Author(s): Mitchell H. Gail* and Yunhu Wan
• Companies: National Cancer Institute and Leidos Corporation
• Keywords: Microbiome; Beta diversity; Power; Standard reference samples; PERMANOVA; MiRKAT

Abstract

Comparisons of microbiome communities are often based on pairwise beta-diversity measures, such as Bray-Curtis dissimilarity. For two-group discrimination, a permutation test (PERMANOVA) and a kernel expansion test (MiRKAT) are used. A recent proposal computes mean dissimilarities from a microbiome test sample to a set of standard reference samples from different sites, producing a vector of mean dissimilarities for each test sample. Standard statistical procedures can be performed on these vectors. We compare the power of Hotelling’s T-squared test applied to reference vectors to that of MiRKAT and PERMANOVA for two-group discrimination using 16S data from the Human Microbiome Project. We also compare the power of regressions based on reference vectors to that of MiRKAT for detecting an association of body mass index with beta-diversity in 16S data from the American Gut Project. Using vectors of mean dissimilarities to standard reference groups can promote data sharing and comparisons of results by different investigators, and these methods have power comparable to MiRKAT, that requires sharing DNA data to construct beta-diversity matrices.