An adaptive microbiome α-diversity-based association analysis method.

ABSTRACT: To relate microbial diversity with various host traits of interest (e.g., phenotypes, clinical interventions, environmental factors) is a critical step for generic assessments about the disparity in human microbiota among different populations. The performance of the current item-by-item α-diversity-based association tests is sensitive to the choice of α-diversity metric and unpredictable due to the unknown nature of the true association. The approach of cherry-picking a test for the smallest p-value or the largest effect size among multiple item-by-item analyses is not even statistically valid due to the inherent multiplicity issue. Investigators have recently introduced microbial community-level association tests while blustering statistical power increase of their proposed methods. Howe

SUBMITTER: Koh H 

PROVIDER: S-EPMC6303306 | biostudies-literature | 2018 Dec

REPOSITORIES: biostudies-literature