Pick features for cell type prediction with caret by LOOCV

Pick features for cell type prediction with caret by LOOCV

Usage

pickCompProbesCaretLOOCV(
  betas,
  meta,
  ct,
  ps,
  min.delta.beta,
  p.val,
  caretMods,
  filterK = 1000,
  seed = 1234,
  plot = TRUE,
  verbose = TRUE
)

Arguments

betas

A beta matrix of reference DNA methylation data that will be used to select for features for cell type prediction

meta

A data frame of phenotype data, specifying which of the reference samples are of what cell types

ct

A vector of characters specifying the cell types to deconvolute

ps

A character of either "any" or "both" to specify first line filter for T test prior to passing to machine learning algorithms

min.delta.beta

A numeric variable defining T test minimum delta beta for first line filter prior to passing to machine learning algorithms

p.val

A numeric variable defining maximum T test p.value for first line filter prior to passing to machine learning algorithms

caretMods

A vector of characters, selecting the models to use to pick the cell type prediction features, options include "lasso", "EL", "BLR", "CART", "RF", "GBM", "PLDA", "GAnRF", "GAnNB", "GAnSVM", and "GAnNN"

filterK

An integer, representing the number probes to input to the machine learning algorithms with top T test probes

seed

An integer specifying the seed for reproducibility

plot

A Boolean specifying whether to plot a heatmap showing the clustering performance using the probes selected

verbose

A Boolean specifying whether the function should be verbose or not