JSM 2011 Online Program

The views expressed here are those of the individual authors and not necessarily those of the JSM sponsors, their officers, or their staff.

Abstract Details

Activity Number: 359
Type: Contributed
Date/Time: Tuesday, August 2, 2011 : 10:30 AM to 12:20 PM
Sponsor: Biometrics Section
Abstract - #302334
Title: Dirichlet-Multinomial Power Calculations and Statistical Tests for Microbiome Data
Author(s): Patricio La Rosa*+ and William Shannon and Elena Deych and George Weinstock and Erica Sodergren and Paul Brooks and Edward Boone and Qin Wang
Companies: Washington University School of Medicine and Washington University School of Medicine and Washington University School of Medicine and Washington University School of Medicine and Washington University School of Medicine and Virginia Commonwealth University and Virginia Commonwealth University and Virginia Commonwealth University
Address: 660 S Euclid Ave, Box 8005, St Louis, MO, 63110,
Keywords: Microbiome ; dirichlet-multinomial ; next-gen sequencing ; power/sample size ; diversity indices ; applied biostatistics
Abstract:

We provide a statistical framework to perform formal hypothesis testing, and to calculate power and sample size requirements for human microbiome experiments using taxonomical classification of metagenomic sequences. The methods proposed allow for modeling and comparing statistically the taxa abundance distributions of microbiotas from one and several populations. In particular, we use the Dirichlet-Multinomial (DM) distribution to model taxa counts from a set of samples. The DM model takes into account the variability of the taxa probabilities or relative abundance distribution (RAD) across samples providing a better fit to the data than a Multinomial model. We study the power and size of the following hypothesis tests: Multinomial goodness of fit against a Dirichlet multinomial alternative; one-sample, two-sample, and multiple-sample comparison of RAD means; and two-sample comparison of RAD distributions. More specifically, for a given taxa number, we provide guidelines for computing sample size, namely, the numbers of subjects and number of reads per subject required to obtain a desired statistical power. As an example, we apply our methodology to the HMP data on 24 subjects.


The address information is for the authors that have a + after their name.
Authors who are presenting talks have a * after their name.

Back to the full JSM 2011 program




2011 JSM Online Program Home

For information, contact jsm@amstat.org or phone (888) 231-3473.

If you have questions about the Continuing Education program, please contact the Education Department.