Skip to contents

Pick features for cell type prediction

pickProbes.Rd

Pick features for cell type prediction

Usage

pickProbes(
  dataNormed,
  probeList = c("Ttest", "Caret", "IDOL", "DHS"),
  probeSelect = c("any", "both"),
  nProbes,
  caretMods = c("lasso", "EL", "BLR", "CART", "RF", "GBM", "GAnLDA", "GAnRF", "GAnNB",
    "GAnSVM", "GAnNN"),
  seed = 1,
  p.val = 0.05,
  min.delta.beta = 0,
  filterK = 1000,
  plotRef = TRUE,
  verbose = TRUE
)

Arguments

dataNormed

A list of dataframe containing the normalized data, output of combData()

probeList

A character, specifying how the probe should be selected. Options include "Ttest", "Caret_CV", "Caret_LOOCV", and "IDOL"

probeSelect

If probeSelect = Ttest, specify how should the top probes be picked from T test output, options include "both", "any", "pval"

nProbes

An integer specifying the number of probes to pick for each cell type

caretMods

If probeSelect %in% c("Caret_CV", "Caret_LOOCV"), input a vector of characters that specify the models to use for feature selection

seed

An integer specifying the seed for reproducibility

p.val

If probeSelect = Ttest, specify a numeric value for maximum pvalue cutoff

min.delta.beta

If probeSelect = Ttest, specify a numeric value for minimum delta beta (effect size) cutoff

filterK

If probeSelect %in% c("Caret_CV", "Caret_LOOCV"), input an integer representing the number probes to input to the machine learning algorithms with top T test probes

plotRef

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

Value

A list containing the features selected by the specified methods, and the coefficients of the selected probes for downstream prediction

Examples