regressFactors: run (e.g.) eBayes(lmFit(assay(x), design)) on columns of...

In ttriche/sesamizeGEO: the poor man's methylation suite (cf. `sesamizer`)

run (e.g.) eBayes(lmFit(assay(x), design)) on columns of (e.g.) assay(x)

Description

if y is specified, use that instead of assay(x) but constrain it the same if purify is specified, purify assay(x) before proceeding with the fit(s)

Usage

regressFactors(
  x,
  columns,
  y = NULL,
  how = c("lmFit", "concordance"),
  check = TRUE,
  purify = FALSE,
  purity = NULL,
  betas = FALSE,
  ...
)

Arguments

x	a SummarizedExperiment-like object
columns	colData(x) column names to regress as predictors
y	optional matrix w/ ncols(x) columns (default is assay(x))
how	"lmFit" (default) or "concordance" (survival::concordance)
check	ensure that the selected columns of y are non-NA? (TRUE)
purify	purify x? (default is FALSE, but this can be useful)
purity	if purify == TRUE, a required vector of purities for x
betas	if purify == TRUE, are the factors beta values? (FALSE)
...	additional arguments to lmFit, if how == "fit"

Value

       a fitted model (use topTable on it) or concordance object

See Also

      purify

      limma::lmFit

      limma::topTable

      survival::concordance

ttriche/sesamizeGEO documentation built on Nov. 12, 2023, 5:42 p.m.