Skip to contents

Performs a Canonical Correspondence Analysis (CCA) or a Redundancy Analysis (RDA) of the abundance table constrained by environmental variables, and plots the samples and the environmental vectors as a biplot.

Usage

cca_rda_biplot(
  table,
  env_data = NULL,
  env_vars,
  method = "hell",
  metadata,
  group_col = NULL,
  group_colors = NULL,
  legend_title = NULL,
  scale_env = TRUE,
  pval_threshold = 0.05,
  p_adjust_method = "none",
  show_all_env_vectors = FALSE,
  analysis = "CCA",
  scale_arrows = 1,
  title = "auto",
  save_table = FALSE,
  table_filename = "cca_rda_scores.txt"
)

Arguments

table

A data frame with taxa in rows and samples in columns. The last column must be named taxonomy, containing full taxonomic strings.

env_data

Optional data frame of environmental variables, with row names matching the sample names in table. Default NULL: the variables named in env_vars are taken from metadata. Use it only when the variables are in a separate table.

env_vars

A character vector with the names of environmental variables (columns of metadata, or of env_data if given) to include in the analysis.

method

Character. Transformation of the abundances, passed to vegan::decostand(). Default "hell" (Hellinger).

metadata

A data frame containing sample metadata. Its first column must hold the sample IDs (the column names of table). The environmental variables can be columns of metadata (see env_vars).

group_col

Character. Name of the column in metadata that defines the groups. Used to color the samples. Optional; NULL (default) for no groups.

group_colors

Optional character vector of colors, one per group, named after the groups or in their order. If NULL (default), the colorblind-friendly Okabe-Ito palette is used.

legend_title

Character. Title of the legend. If NULL (default), the name of group_col is used.

scale_env

Logical. If TRUE (default), the environmental variables are scaled (without centering) before the analysis.

pval_threshold

Numeric. P-value cutoff (after p_adjust_method) to draw only the significant environmental variables. Default 0.05.

p_adjust_method

Character. Multiple-testing correction for the vegan::envfit() p-values of the environmental variables, any method of stats::p.adjust(). Default "none" (raw p-values), the usual choice with few variables.

show_all_env_vectors

Logical. If TRUE, all the environmental variables are drawn, significant or not. Default FALSE.

analysis

Constrained ordination method: "CCA" (default, Canonical Correspondence Analysis) or "RDA" (Redundancy Analysis). Case-insensitive.

scale_arrows

Numeric. Multiplies the length of the environmental arrows; it only changes how they are drawn. Default 1.

title

Character. Plot title. "auto" (default) shows "CCA Biplot" or "RDA Biplot"; NULL shows no title; any other text is used as the title.

save_table

Logical. If TRUE, saves the sample scores and environmental vector loadings (one table, with a type column: "site" or "vector") as a tab-delimited file. Default FALSE.

table_filename

Character. Name or path of the saved file (used when save_table = TRUE). Default "cca_rda_scores.txt".

Value

A ggplot object with the biplot. The axis titles show the percentage of the variance (inertia) explained by each constrained axis.

Details

The p-values of the environmental vectors come from vegan::envfit() permutations; call set.seed() before the function to make them reproducible.

Examples

table_path <- system.file("extdata", "subset_table_bacteria.txt", package = "MicroBioMeta")
table <- read.delim(table_path, row.names = 1, check.names = FALSE)

metadata_path <- system.file("extdata", "subset_metadata_bacteria.txt", package = "MicroBioMeta")
metadata <- read.delim(metadata_path, check.names = FALSE)

# The environmental variables are columns of metadata, chosen with env_vars
cca_rda_biplot(
    table                = table,
    metadata             = metadata,
    env_vars             = c("pH", "TOC", "FW", "Root_FW", "DW"),
    analysis             = "RDA",
    show_all_env_vectors = TRUE,
    group_col            = "Location"
)