MIMIX (MIcrobiome MIXed model) is hierarchical Bayesian model for the analysis of next-generation sequencing OTU abundance data from designed experiments. It achieves four scientific objectives:
- Global tests of whether experimental treatments affect microbiome composition,
- Local tests for treatment effects on individual taxa and estimation of theses effects if present,
- Quantification of how different sources of variability contribute to microbiome heterogeneity, and
- Characterization of latent structure in the microbiome, which may suggest ecological subcommunities.
For more information, please see our open access e-print: MIMIX: a Bayesian Mixed-Effects Model for Microbiome Data from Designed Experiments.
Description of files
analyze.jl runs the full-scale NutNet data analysis with MIMIX, MIMIX w/o Factors, and PERMANOVA w/ Bray-Curtis.
data contains the NutNet data stored as
Y.csv (sequence counts),
X.csv (experimental treatments), and
Z.csv (blocking factors).
demo.ipynb demonstrates fitting MIMIX to a subset of the real data.
models.jl defines the model hierarchy and posterior sampling schemes for MIMIX and MIMIX w/o Factors.
results contains three subdirectories,
validate, which store the output from their respective scripts.
simulate.jl runs the simulation study comparing MIMIX, MIMIX w/o Factors, and PERMANOVA w/ Bray-Curtis.
utils.jl defines helper functions for the simulation study and cross-validation files.
validate.jl runs a five-fold cross-validation to compare the fits of MIMIX and MIMIX w/o Factors to the real data.