GitHub - waldronlab/bugSigSimple: Simple exploratory analysis of curated microbe signatures (original) (raw)

bugSigSimple: analyses of bugsigdb.org

This repo demonstrates several analyses of curated microbe signatures fromhttps://bugsigdb.org. See "Articles" for existing analyses, and "Reference" for functions provided by this packages.

Installation

bugSigSimple requires Bioconductor version 3.14 or later:

if (!require("BiocManager", quietly = TRUE)) install.packages("BiocManager") BiocManager::install() BiocManager::version()

To install this bugSigSimple and use these functions, do:

BiocManager::install("remotes", dependencies = TRUE) BiocManager::install("waldronlab/bugSigSimple", dependencies = TRUE)

  1. Delivery Mode and the Meconium Microbiome by Shaimaa Elsafoury for a basic creation of a study table and tables of the frequency of the most commonly identified taxa.
  2. The COVID-19 associated microbiome by Clare Grieve: for professional tables, clustered heatmaps and interactive heatmaps for identifying similar signatures, and hierarchical dendrograms.
  3. An analysis of Most frequent taxa in Major Depression and Bipolar Disorder by Fatima Azhar: This analysis does not take advantage of current functionality provided by this package and by BugSigDBStats, but it demonstrates a manual binomial test that can be used to test H_0H_0H_0: a taxon is equally probable to be reported as increased abundance or decreased abundance. This is only a sufficiently powered test when the same taxon is reported in at least 6 different studies (binom.test(0, 6) -> p=0.03125))
  4. The Irritable Bowel Syndrome-associated Microbiome by Kweku Amoo: Similar to 1 in another context.
  5. The Endometriosis-associated Microbiome by Samara Khan: This analysis generates a hypothesis based on frequently identified genera in bugsigdb.org, then tests the hypothesis with individual-participant data from curatedMetagenomicData, using linear regression, t-test, a Poisson model, and (zero-inflated) negative binomial log-linear models.
  6. Analysis of prevalent stool genera in bugsigdb.org by Haoyan Zhong: This analysis utilizes a set of signatures of prevalent species and genera found in healthy human microbiomes (https://bugsigdb.org/Study_562) for overlap with other signatures in bugsigdb.org. It focuses on gut microbiome signatures, but there are signatures of healthy prevalent genera and species available for other body sites as well.