simple_varpart: Use variancePartition to try and understand where the...

In elsayed-lab/hpgltools: A pile of (hopefully) useful R functions

Use variancePartition to try and understand where the variance lies in a data set.

Description

The arguments and usage of variancePartition are a bit opaque. This function attempts to fill in reasonable values and simplify its invocation.

Usage

simple_varpart(
  expt,
  predictor = NULL,
  factors = c("condition", "batch"),
  chosen_factor = "batch",
  do_fit = FALSE,
  cor_gene = 1,
  cpus = NULL,
  genes = 40,
  parallel = TRUE,
  strict_filter = TRUE,
  mixed = FALSE,
  modify_expt = TRUE
)

Arguments

expt	Some data
predictor	Non-categorical predictor factor with which to begin the model.
factors	Character list of columns in the experiment design to query
chosen_factor	When checking for sane 'batches', what column to extract from the design?
do_fit	Perform a fitting using variancePartition?
cor_gene	Provide a set of genes to look at the correlations, defaults to the first gene.
cpus	Number cpus to use
genes	Number of genes to count.
parallel	Use doParallel?
strict_filter	Perform a strict filtering of the results via median_by_factor and dropping any genes with a 0.
mixed	Used a mixed model?
modify_expt	Add annotation columns with the variance/factor?

Value

List of plots and variance data frames

See Also

[variancePartition] DOI:10.1186/s12859-016-1323-z.

elsayed-lab/hpgltools documentation built on May 9, 2024, 5:02 a.m.