amp_diffabund: Differential abundance test

In KasperSkytte/ampvis2extras: Extra (experimental) functions for the ampvis2 package

Description

Tests if there is a significant difference in abundances between samples or groups hereof based on selected conditions. Returns a list containing test results as well as two different plots; an MA-plot and an abundance plot with taxa with the most significant p-value (below the threshold).

Usage

amp_diffabund(data, group)

Arguments

data	(required) Data list as loaded with amp_load().
group	(required) A categorical variable in the metadata that defines the sample groups to test.
test	The name of the test to use, either "Wald} or \code{"LRT. See DESeq. (default: "Wald")
fitType	The type of fitting of dispersions to the mean intensity, either "parametric", "local", or "mean". (default: "parametric")
num_threads	The number of threads to use for parallelization by the BiocParallel backend. Parallelization is not supported on windows machines. (default: 1)
signif_thrh	Significance threshold. (default: 0.01)
fold	Log2fold filter for displaying significant results. (default: 0)
verbose	(Logical) Whether to print status messages during the test calculations. (Default: TRUE)
signif_plot_type	Either "boxplot" or "point". (default: "point")
plot_nshow	The amount of the most significant results to display in the most-significant plot. (default: 10)
plot_point_size	The size of the plotted points. (default: 2)
tax_aggregate	The taxonomic level to aggregate the OTUs. (default: "OTU")
tax_add	Additional taxonomic level(s) to display, e.g. "Phylum". (default: NULL)
tax_class	Converts a specific phylum to class level instead, e.g. "p__Proteobacteria".
tax_empty	How to show OTUs without taxonomic information. One of the following: "remove": Remove OTUs without taxonomic information. "best": (default) Use the best classification possible. "OTU": Display the OTU name.
adjust_zero	Keep 0 abundances in ggplot2 median calculations by adding a small constant to these.
...	Additional arguments passed on to DESeq.

Value

A list with multiple elements:

• "DESeq2_results": The raw output result from DESeq.

• "DESeq2_results_signif": The raw output result from DESeq, but subset to only taxa with p-value below the threshold set by signif_thrh.

• "signif_plotdata": The data used to generate the ggplots, but subset to only taxa with p-value below the threshold set by signif_thrh.

• "Clean_results": A simpler version of DESeq2_results_signif only with adjusted p-values, log2FoldChange, and average abundance of each taxa per group.

• "plot_MA": MA-plot

• "plot_MA_plotly": Interactive plotly plot of MA-plot with custom hover information.

• "plot_signif": Abundance plot with taxa with the n most significant p-value (below the threshold), where n is set by plot_nshow.

• "plot_signif_plotly": Interactive plotly plot of plot_signif with custom hover information.

Author(s)

Kasper Skytte Andersen [e-mail]

Mads Albertsen [e-mail]

Examples

library(ampvis2extras)
# Load example data
data("AalborgWWTPs")

# Subset to a few taxa, save the results in an object
d <- amp_subset_taxa(AalborgWWTPs, tax_vector = c("p__Chloroflexi", "p__Actinobacteria"))
results <- amp_diffabund(d, group = "Plant", tax_aggregate = "Genus")

# Show plots
results$plot_signif
results$plot_MA

# Or show raw results
results$Clean_results

KasperSkytte/ampvis2extras documentation built on Oct. 3, 2021, 11:38 a.m.