Perform differential analysis using edgeR

Source: R/DA-edgeR.R

Differential expression analysis based on the Negative Binomial distribution using edgeR.

run_edger(
  ps,
  group,
  confounders = character(0),
  contrast = NULL,
  taxa_rank = "all",
  method = c("LRT", "QLFT"),
  transform = c("identity", "log10", "log10p"),
  norm = "TMM",
  norm_para = list(),
  disp_para = list(),
  p_adjust = c("none", "fdr", "bonferroni", "holm", "hochberg", "hommel", "BH", "BY"),
  pvalue_cutoff = 0.05,
  ...
)

Arguments

ps

ps a phyloseq::phyloseq object.

group

character, the variable to set the group, must be one of the var of the sample metadata.

confounders

character vector, the confounding variables to be adjusted. default character(0), indicating no confounding variable.

contrast

this parameter only used for two groups comparison while there are multiple groups. For more please see the following details.

taxa_rank

character to specify taxonomic rank to perform differential analysis on. Should be one of phyloseq::rank_names(phyloseq), or "all" means to summarize the taxa by the top taxa ranks (summarize_taxa(ps, level = rank_names(ps)[1])), or "none" means perform differential analysis on the original taxa (taxa_names(phyloseq), e.g., OTU or ASV).

method

character, used for differential analysis, please see details below for more info.

transform

character, the methods used to transform the microbial abundance. See transform_abundances() for more details. The options include:

  • "identity", return the original data without any transformation (default).

  • "log10", the transformation is log10(object), and if the data contains zeros the transformation is log10(1 + object).

  • "log10p", the transformation is log10(1 + object).

norm

the methods used to normalize the microbial abundance data. See normalize() for more details. Options include:

  • "none": do not normalize.

  • "rarefy": random subsampling counts to the smallest library size in the data set.

  • "TSS": total sum scaling, also referred to as "relative abundance", the abundances were normalized by dividing the corresponding sample library size.

  • "TMM": trimmed mean of m-values. First, a sample is chosen as reference. The scaling factor is then derived using a weighted trimmed mean over the differences of the log-transformed gene-count fold-change between the sample and the reference.

  • "RLE", relative log expression, RLE uses a pseudo-reference calculated using the geometric mean of the gene-specific abundances over all samples. The scaling factors are then calculated as the median of the gene counts ratios between the samples and the reference.

  • "CSS": cumulative sum scaling, calculates scaling factors as the cumulative sum of gene abundances up to a data-derived threshold.

  • "CLR": centered log-ratio normalization.

  • "CPM": pre-sample normalization of the sum of the values to 1e+06.

norm_para

arguments passed to specific normalization methods. Most users will not need to pass any additional arguments here.

disp_para

additional arguments passed to edgeR::estimateDisp() used for dispersions estimation. Most users will not need to pass any additional arguments here.

p_adjust

method for multiple test correction, default none, for more details see stats::p.adjust.

pvalue_cutoff

numeric, p value cutoff, default 0.05

...

extra arguments passed to the model. See edgeR::glmQLFit() and edgeR::glmFit() for more details.

Value

a microbiomeMarker object.

Details

Note that edgeR is designed to work with actual counts. This means that transformation is not required in any way before inputting them to edgeR.

There are two test methods for differential analysis in edgeR, likelihood ratio test (LRT) and quasi-likelihood F-test (QLFT). The QLFT method is recommended as it allows stricter error rate control by accounting for the uncertainty in dispersion estimation.

contrast must be a two length character or NULL (default). It is only required to set manually for two groups comparison when there are multiple groups. The order determines the direction of comparison, the first element is used to specify the reference group (control). This means that, the first element is the denominator for the fold change, and the second element is used as baseline (numerator for fold change). Otherwise, users do required to concern this parameter (set as default NULL), and if there are two groups, the first level of groups will set as the reference group; if there are multiple groups, it will perform an ANOVA-like testing to find markers which difference in any of the groups.

References

Robinson, Mark D., and Alicia Oshlack. "A scaling normalization method for differential expression analysis of RNA-seq data." Genome biology 11.3 (2010): 1-9.

Robinson, Mark D., Davis J. McCarthy, and Gordon K. Smyth. "edgeR: a Bioconductor package for differential expression analysis of digital gene expression data." Bioinformatics 26.1 (2010): 139-140.

Author

Yang Cao

Examples

data(enterotypes_arumugam)
ps <- phyloseq::subset_samples(
    enterotypes_arumugam,
    Enterotype %in% c("Enterotype 3", "Enterotype 2")
)
run_edger(ps, group = "Enterotype")
#> Warning: Some counts are non-integers, they are rounded to integers.
#> Raw count is recommended for reliable results for edger method.
#> microbiomeMarker-class inherited from phyloseq-class
#> normalization method:              [ TMM ]
#> microbiome marker identity method: [ edgeR: LRT ]
#> marker_table() Marker Table:       [ 34 microbiome markers with 5 variables ]
#> otu_table()    OTU Table:          [ 235 taxa and  24 samples ]
#> sample_data()  Sample Data:        [ 24 samples by  10 sample variables ]
#> tax_table()    Taxonomy Table:     [ 235 taxa by 1 taxonomic ranks ]