R scripts for Human Microbiome Analysis Toolset

A variety of statistical analysis methods to analyze microbiome data are provided in an R script, with formatted example input files. These methods include read count rarefaction, diversity calculations, NMDS clustering, Dirichlet Multinomial Mixtures (DMM) clustering, cluster validation with Silhouette scores and Prediction strength, random forest, and ANCOM-BC2. The script is a valuable tool for microbiome researchers and provides simple-to-use code with comments and properly formatted example input files, to perform a comprehensive set of methods for analyzing microbiome data. Version 2 adds code to filter out taxa that are detected in < 3 samples for the methods where that is appropriate. Version 3 cleans up some redundant code.