Skip to contents

DESeq2 analysis

Source: R/deseq2.R
step_deseq.Rd

Differential expression analysis based on the Negative Binomial (a.k.a. Gamma-Poisson) distribution. This function performs a default analysis through the steps: 1) estimation of size factors: estimateSizeFactors. 2) estimation of dispersion: estimateDispersions. 3) Negative Binomial GLM fitting and Wald statistics: nbinomWaldTest. For complete details on each step, see the manual pages of the respective functions. After the DESeq function returns a DESeqDataSet object, results tables (log2 fold changes and p-values) can be generated using the results function. Shrunken LFC can then be generated using the lfcShrink function.

Usage

step_deseq(
  rec,
  test = "Wald",
  fitType = "local",
  betaPrior = FALSE,
  type = "ashr",
  max_significance = 0.05,
  log2FC = 0,
  rarefy = FALSE,
  id = rand_id("deseq"),
  engine_args = list()
)

Arguments

rec

A Recipe object. The step will be added to the sequence of operations for this Recipe. Execution requires add_model() before prep(); the step can be configured before adding the model.

test

Either "Wald" or "LRT", which will then use either Wald significance tests (defined by nbinomWaldTest), or the likelihood ratio test on the difference in deviance between a full and reduced model formula (defined by nbinomLRT).

fitType

either "parametric", "local", "mean", or "glmGamPoi" for the type of fitting of dispersions to the mean intensity. See estimateDispersions for description.

betaPrior

whether or not to put a zero-mean normal prior on the non-intercept coefficients See nbinomWaldTest for description of the calculation of the beta prior. In versions >=1.16, the default is set to FALSE, and shrunken LFCs are obtained afterwards using lfcShrink.

type

"apeglm" is the adaptive Student's t prior shrinkage estimator from the 'apeglm' package; "ashr" is the adaptive shrinkage estimator from the 'ashr' package, using a fitted mixture of normals prior - see the Stephens (2016) reference below for citation; "normal" is the 2014 DESeq2 shrinkage estimator using a Normal prior.

max_significance

The q-value threshold for significance.

log2FC

log2FC cutoff.

rarefy

Boolean indicating if OTU counts must be rarefyed. This rarefaction uses the standard R sample function to resample from the abundance values in the otu_table component of the first argument, physeq. Often one of the major goals of this procedure is to achieve parity in total number of counts between samples, as an alternative to other formal normalization procedures, which is why a single value for the sample.size is expected. If 'no_seed', rarefaction is performed without a set seed.

id

Unique, non-empty identifier for this step. DA step IDs must not use the reserved internal names taxa_id, contrast_id, comparison, contrast_type, var, effect, sum, .tax_order, .contrast_order or .effect_order. Other identifiers, including names such as .custom, are retained unchanged.

engine_args

Named lists of advanced arguments for the native size_factors, fit, results, or shrink stage. Arguments managed by dar or exposed above cannot be overridden.

Value

An object of class Recipe

See also

Other Diff taxa steps: step_aldex2(), step_aldex3(), step_ancom(), step_corncob(), step_lefse(), step_linda(), step_maaslin(), step_rademu(), step_wilcox()

Examples

data(metaHIV_phy)

rec <- recipe(metaHIV_phy) |>
  add_model(~ RiskGroup2, targets = "RiskGroup2", tax_level = "Phylum")

rec_default <- step_deseq(rec)

# Pass advanced arguments to native DESeq2 stages.
rec_custom <- step_deseq(
  rec,
  engine_args = list(
    fit = list(minReplicatesForReplace = Inf),
    shrink = list(lfcThreshold = 1)
  )
)