Combining p-values from various statistical methods for microbiome data
Authors:
Open Access
- 10 November 2022
- journal article
- research article
- Published by Frontiers Media SA in Frontiers in Microbiology
- Vol. 13
Abstract
Motivation: In the field of microbiome analysis, there exist various statistical methods that have been developed for identifying differentially expressed features, that account for the overdispersion and the high sparsity of microbiome data. However, due to the differences in statistical models or test formulations, it is quite often to have inconsistent significance results across statistical methods, that makes it difficult to determine the importance of microbiome taxa. Thus, it is practically important to have the integration of the result from all statistical methods to determine the importance of microbiome taxa. A standard meta-analysis is a powerful tool for integrative analysis and it provides a summary measure by combining p-values from various statistical methods. While there are many meta-analyses available, it is not easy to choose the best meta-analysis that is the most suitable for microbiome data.Results: In this study, we investigated which meta-analysis method most adequately represents the importance of microbiome taxa. We considered Fisher’s method, minimum value of p method, Simes method, Stouffer’s method, Kost method, and Cauchy combination test. Through simulation studies, we showed that Cauchy combination test provides the best combined value of p in the sense that it performed the best among the examined methods while controlling the type 1 error rates. Furthermore, it produced high rank similarity with the true ranks. Through the real data application of colorectal cancer microbiome data, we demonstrated that the most highly ranked microbiome taxa by Cauchy combination test have been reported to be associated with colorectal cancer.Keywords
This publication has 69 references indexed in Scilit:
- Differential abundance analysis for microbial marker-gene surveysNature Methods, 2013
- Development of a Dual-Index Sequencing Strategy and Curation Pipeline for Analyzing Amplicon Sequence Data on the MiSeq Illumina Sequencing PlatformApplied and Environmental Microbiology, 2013
- Variable selection for sparse Dirichlet-multinomial regression with an application to microbiome data analysisThe Annals of Applied Statistics, 2013
- Quality-filtering vastly improves diversity estimates from Illumina amplicon sequencingNature Methods, 2013
- Human Intestinal Lumen and Mucosa-Associated Microbiota in Patients with Colorectal CancerPLOS ONE, 2012
- Introducing EzTaxon-e: a prokaryotic 16S rRNA gene sequence database with phylotypes that represent uncultured speciesInternational Journal of Systematic and Evolutionary Microbiology, 2012
- edgeR: a Bioconductor package for differential expression analysis of digital gene expression dataBioinformatics, 2009
- Combining dependent P-valuesStatistics & Probability Letters, 2002
- An improved Bonferroni procedure for multiple tests of significanceBiometrika, 1986
- The American Soldier. Vol. I, Adjustment During Army Life; The American soldier. Vol. II, Combat and Its Aftermath.Journal of Consulting Psychology, 1949