Compositional_data_microbiome_statistics

CODA

Scripts and concepts for statistical analysis of microbiome data. Important to note that microbiome data is compositional in nature.

1. Normalisation methods-

CLR (Centred Log Ratio)- available tools and packages

• https://github.com/maxibor/microbiomeNormalization (package- compositions)
• https://cran.r-project.org/web/packages/Hotelling/Hotelling.pdf (package- hotelling)
• https://search.r-project.org/CRAN/refmans/compositions/html/clr.html
• https://cran.r-project.org/web/packages/compositions/index.html

Deal with zero count values in sparse data-

• zcompositions R package
• ALDEX2 (Bioconductor)- includes clr and it deals with compositional data using log-ratio transformation