BayesianNetworkRegression.jl

Author solislemuslab
Popularity
9 Stars
Updated Last
12 Months Ago
Started In
June 2021

Bayesian Network Regression model for microbiome networks as covariates on biological phenotypes as response bayesiannetworkregression logo

CI codecov

Overview

BayesianNetworkRegression.jl is a Julia package to perform (Bayesian) statistical inference of a regression model with networked covariates on a real response.

The structure of the data assumes $n$ samples, each sample with a microbial network represented as an adjacency matrix $A_i$ and a measured biological phenotype $y_i$. This package will identify influential nodes (microbes) associated with the phenotype, as well as significant edges (interactions among microbes) that are also associated with the phenotype.

Examples of data that can be fit with this model:

  • Samples of soil microbiome data represented as $n$ microbial networks and measured yield of $n$ plants on that soil. The model will identify microbes and interactions associated with plant yield.
  • Samples of gut microbiome data represented as $n$ microbial networks, one per human patient, and BMI measurement per human patient. The model will identify microbes and interactions associated with BMI.

Usage

BayesianNetworkRegression.jl is a julia package, so the user needs to install julia, and then install the package.

To install the package, type inside Julia:

]
add BayesianNetworkRegression

Help and errors

To get help, check the documentation here. Please report any bugs and errors by opening an issue.

Citation

If you use BayesianNetworkRegression.jl in your work, we kindly ask that you cite the following paper:

@article{Ozminkowski2022,
author = {Ozminkowski, S. and Sol'{i}s-Lemus, C.},
year = {2022},
title = {{Identifying microbial drivers in biological phenotypes with a Bayesian Network Regression model}},
journal = {arXiv:2208.05600}
}

License

BayesianNetworkRegression.jl is licensed under a GNU General Public License v2.0.

Contributions

Users interested in expanding functionalities in BayesianNetworkRegression.jl are welcome to do so. See details on how to contribute in CONTRIBUTING.md.