BIRDMAn: Bayesian Inferential Regression for Differential Microbiome Analysis

BIRDMAn is a framework for performing differential abundance on microbiome data through a Bayesian lens. We provide several default models but also allow users to create their own statistical models based on their specific experimental designs/questions.

Installation

Note

BIRDMAn requires Python >= 3.8

There are several dependencies you must install to use BIRDMAn. The easiest way to install the required dependencies is through conda or mamba.

mamba install -c conda-forge biom-format patsy xarray arviz cmdstanpy
pip install birdman

If you are planning on contributing to BIRDMAn you must also install the following packages:

• pytest

• scikit-bio

• sphinx

Docker

We have also provided a Docker image that you can use for running BIRDMAn.

User Guide

• Installing cmdstanpy
• Implementing a default model
• Implementing a custom model
  • Downloading data from Qiita
  • Processing metadata
  • Filtering feature table
  • Model specification
  • Stan code
  • Running BIRDMAn
  • Converting to InferenceData
• Diagnosing BIRDMAn models
  • Rhat
  • PSIS-LOO-CV
  • Posterior predictive check
  • Comparing to null model
• Writing Stan code
  • Structure of a Stan program
  • Data
  • Parameters
  • Transformed parameters
  • Model
  • Generated quantities
• Working with arviz
  • xarray
  • arviz
• Parallelization in BIRDMAn through the Single Feature Model
  • Stan code
  • Getting started
  • Define SingleFeatureModel class
  • Chunking
  • Running our parallel script

How BIRDMAn Works

• Bayesian inference
  • Markov Chain Monte Carlo
  • Sampling from the posterior
  • Resources
• Negative binomial models
  • Count models
  • Negative binomial model
  • Output

API

• BIRDMAn Model API
  • Default Models
  • Table Model
  • Single Feature Model
  • Model Iterator
  • Base Model
• Inference Functions
  • concatenate_inferences()
  • fit_to_inference()
  • stan_var_to_da()
• Diagnostics Functions
  • ess()
  • loo()
  • r2_score()
  • rhat()
• Summary Functions
  • summarize_posterior()
• Transformation Functions
  • posterior_alr_to_clr()
• Stats Functions
  • hotelling_ttest()
• Visualization Functions
  • plot_parameter_estimates()
  • plot_posterior_predictive_checks()

Indices and tables

• Index

• Module Index

• Search Page