variable_correlation_network: Create a node network ggplot object of the correlation of...
In schuyler-smith/phylosmith: Functions to help analyze data as phyloseq objects

View source: R/variable_correlation_network.R

variable_correlation_network

R Documentation

Create a node network ggplot object of the correlation of taxa and sample variables from a phyloseq object. Function from the phylosmith-package.

Description

Inputs a phyloseq-class object and creates a network from the co-occurrence. The co-occurrence can either be input, or it will be calculated with the Spearman-rank correlation. Also, the layout of the graph can be given as an argument as well for reproducibility.

Usage

variable_correlation_network(phyloseq_obj, variables, classification = NULL,
treatment = NULL, subset = NULL, correlation_table = NULL, method = 'spearman',
rho_threshold = c(-0.01, 0.01), p_threshold = 0.05, colors = 'default',
negative_positive_colors = c('tomato3', 'gray22'))

Arguments

`phyloseq_obj`	A `phyloseq-class` object. It must contain `sample_data()`) with information about each sample, and it must contain `tax_table()`) with information about each taxa/gene.
`variables`	Numerical factors within the in the `sample_data` to correlate with the abundance data.
`classification`	Column name as a string or number in the `tax_table` for the factor to use for node colors.
`treatment`	Column name as a string or number in the `sample_data`. This can be a vector of multiple columns and they will be combined into a new column.
`subset`	A factor within the `treatment`. This will remove any samples that to not contain this factor. This can be a vector of multiple factors to subset on.
`correlation_table`	Table of the correlation of taxa/variables in the `phyloseq_obj`, computed using `variable_correlation`. If no table is given, it will be computed with the `phyloseq_obj`, using the given `treatment` and `p` = 0.05.
`method`	Which correlation method to calculate, "pearson", "spearman", "kendall.
`rho_threshold`	Cutoffs to use to subset the 'correlation_table' by correlation values.
`p_threshold`	Cutoffs to use to subset the 'correlation_table' by singnificance values.
`colors`	Name of a color set from the #' RColorBrewer package or a vector palette of R accepted colors.
`negative_positive_colors`	colors to use for the edges to represent negative and positive correlations.