Factorization of microbiome datasets

This repo was created to exemplify how to use two factorization methods on microbiome datasets. Also, to generate a Random Forest model from them. The codes of these examples were generated during my lab rotation at Rob Knight's Lab - Spring 2019.

Here, you can find examples of using PhyloFactor and DEICODE to analyze a previously published study.

Requirements

• Follow this guide about installing Qiime2 (miniconda or anaconda are required).
• Follow this guide about installing PhyloFactor and an environment for R. For more info, check the original repo and this tutorial.
• Follow the original instructions about installing DEICODE.
• Follow the original instructions about installing Qurro.

Examples

• The datasets used in the examples are available in the data folder.
• Data preprocessing used before the factorization methods is available in this notebook.
• To execute PhyloFactor and generate predictive models, see examples in the notebooks folder.
• To execute DEICODE + Qurro, follow these instructions.