Home

This wiki describes the end-to-end pipeline for analyzing microbial association networks across farms. The analysis consists of two major components:

Between-farms prediction: Using singular value decomposition (SVD) and phylogenetic imputation to predict microbial co-occurrence in one farm based on microbial co-occurrence observed in another.

Permutation test: Assessing the significance of the observed prediction accuracy via link-shuffling permutations.

Each step is automated via R scripts and HPC job wrappers. Follow the sections below to understand and run the full workflow.