treePValues: Create tree of p-values for phyloseq data
In structSSI: Multiple Testing for Hypotheses with Hierarchical or Group Structure

Description Usage Arguments Value Author(s) Examples

This helper function is used to aggregate abundances of individual microbes to higher levels in the tree and test whether those aggregated abundances are significantly different between environments, using the data structures from the phyloseq package framework.

1	treePValues(tree, abundances, environments)

`tree`	An edgelist for a tree containing the phylogenetic relationships between different microbes.
`abundances`	A phyloseq class OTU table specifying the abundances of different microbes across environments.
`environments`	A phyloseq class Sample Data object associating the different environments with variables of interest.

A vector containing the p-values of the linear model predicting the abundances of microbes aggregated to different levels in the taxonomy from environmental variables.

Kris Sankaran

library('igraph')

# Example with random data
if(require('ape')) {
  rand.tree <- as.igraph(rtree(50))
  V(rand.tree)$name <- paste("hyp", 1:50)
  rand.tree <- get.edgelist(rand.tree)
  X <- matrix(rnorm(50 * 4), 50 , 4)
  rownames(X)  <- paste("hyp" , 1:50)
  colnames(X)  <- 1:4
  X[, 1:2] <- X[, 1:2] + 1
  groups <- factor(c("A", "A", "B", "B"))
  treePValues(rand.tree, X, groups)
}

# Example using phyloseq
if(require('ape') & require('phyloseq')) {
    data(chlamydiae)
    abundances <- otu_table(chlamydiae)
    environments <- sample_data(chlamydiae)$SampleType
    ch.tree <- get.edgelist(as.igraph(phy_tree(chlamydiae)))
    ch.pval <- treePValues(ch.tree, abundances, environments)
}