Validate inputs to fit_bread() — validate_bread

Cheap structural checks run before any modeling. Verifies that se is a SummarizedExperiment::SummarizedExperiment with row-level genomic coordinates, that features is a non-empty GenomicRanges::GRanges, that design is a one-sided formula whose variables all live in colData(se), that assay_name is present, and that contrast (when non-NULL) resolves to a coefficient of the design's model matrix.

Usage

validate_bread_input(
  se,
  features,
  design,
  contrast = NULL,
  assay_name = "M",
  input_scale = c("M", "Beta")
)

Arguments

se: A SummarizedExperiment::SummarizedExperiment.
features: A GenomicRanges::GRanges of regions.
design: A one-sided formula, e.g. ~ group + sex.
contrast: Character coefficient name, or NULL to defer.
assay_name: Character scalar naming an assay in se.
input_scale: "M" or "Beta".

Value

Invisibly TRUE. Errors loudly on the first violation.