Compositional transformations can reasonably introduce phenotype-associated values in sparse features

This repository contains the code supporting our work from the manuscript "Compositional transformations can reasonably introduce phenotype-associated values in sparse features". The notebooks contain the works relevant to 1) a synthetic dataset, 2) the vaginal microbiome's association with preterm birth, and 3) The Cancer Genome Atlas (TCGA).

To run the vaginal microbiome analysis, one first needs to download the vaginal microbiome tables from https://www.pnas.org/doi/full/10.1073/pnas.1502875112 under the 'Supporting information' section, specifically the two .csv and .txt files, and place them in the same directory as the notebook.

The tables required to run the TCGA notebooks can be found under the original Poore et al. publication at ftp://ftp.microbio.me/pub/cancer_microbiome_analysis/.

The analyses from all notebooks can complete on a typical laptop in less than 30 seconds.