train_wbcc: Train a predictor for cell-type composition

In molepi/DNAmArray: Streamlined workflow for the quality control, normalization and bias-free analysis of Illumina methylation array data - The Leiden approach

Train a predictor for cell-type composition

Description

Train the cell-type composition predictor using partial least squares regression on mehtylation values

Usage

train_wbcc(
  data,
  covar,
  cellPer,
  model = formula(cellPer ~ covar + data),
  ncomp = 50,
  keep.model = FALSE,
  ...
)

Arguments

data	Matrix with DNA methylation beta-values
covar	Matrix of covariates to correct for in the model
cellPer	Matrix of cell percentages on which the predictor will be trained
model	Formula (default: cellPer ~ covar + data)
ncomp	Number of PLS components to train (default: 50)
keep.model	Logical specifying whether to return the model (default: FALSE)
...	Additional parameters for plsr see ?plsr

Value

Prediction model PLSR object

Author(s)

mvaniterson

molepi/DNAmArray documentation built on Aug. 24, 2022, 1:42 p.m.