Enrich BREAD classifications against KnowYourCG databases

For each region classified as "hypermethylated" or "hypomethylated" (or any level you pick), extracts the constituent EPIC / HM450 / MM285 probes and runs knowYourCG::testEnrichment() against one or more CpG annotation databases. This is how you get from a region-level BREAD result to "is the hyper set enriched for Polycomb targets / AP-1 binding sites / PMDs / ...".

Usage

bread_kycg(
  fit,
  which = c("hypermethylated", "hypomethylated"),
  databases = NULL,
  platform = c("EPIC", "EPICv2", "HM450", "MM285"),
  universe = NULL,
  alternative = "greater",
  include_genes = FALSE
)

Arguments

fit

A BreadFit.

which

Classification level(s) to enrich. Default c("hypermethylated", "hypomethylated") runs both as separate queries. Use a single string to run one.

databases

Character vector of KYCG database group titles. Default NULL selects defaults by platform.

platform

One of "EPIC", "EPICv2", "HM450", "MM285". Default "EPIC". Only used when databases = NULL.

universe

Optional universe of probe IDs (character vector). Default NULL uses all probes in the BREAD mapping (i.e. probes covered by features), which is the right universe for a region-level enrichment.

alternative

"greater" (default), "two.sided", or "less".

include_genes

Passed through to knowYourCG::testEnrichment().

Value

A tidy data.frame with one row per tested set, containing a query column naming the classification level ("hypermethylated" / "hypomethylated" / ...) alongside standard KYCG columns (dbname, estimate, p.value, FDR, ...).

Databases

If databases = NULL, the function picks a sensible default set for the platform you specify via platform (e.g. "EPIC", "EPICv2", "HM450", "MM285"). Pass a character vector of knowYourCG::listDBGroups() titles to override.

Examples

if (FALSE) { # \dontrun{
  fit <- fit_bread(se, features, ~ group)
  enr <- bread_kycg(fit, platform = "EPIC")
  head(enr[enr$FDR < 0.01, ])
} # }