Test differential exon or transcript usage from a fitted limma/edgeR model

Usage

dtu_diffsplice(
  fit,
  gene_id,
  feature_id = NULL,
  robust = FALSE,
  coef = NULL,
  contrast = NULL,
  ...
)

Arguments

fit: A native MArrayLM or DGEGLM model fit.
gene_id: Gene identifier for every fitted feature.
feature_id: Optional exon or transcript identifier.
robust: Use robust empirical Bayes estimation.
coef, contrast: Coefficient or contrast for the edgeR method. Supply at most one. For limma, apply contrasts to fit before calling this function; limma's diffSplice() calculates statistics for the fitted coefficients.
...: Additional arguments passed to the selected backend's diffSplice() method.

Value

A native limma or edgeR differential-splicing result.

Details

gene_id and feature_id are aligned to the fitted rows when they are named. This is particularly useful after expression filtering. Supply an untested DGEGLM to the edgeR method; choose its test with coef or contrast. For limma, apply the desired contrast with limma::contrasts.fit() before calling this function when a contrast, rather than a coefficient, is required.