README.md

wbccPredictor

Whole blood cell count predictor

Packages for using partial least squares regression (PLSR) to estimate cell count proportions in a sample with methylation data (Houseman et al, 2012). This package is based on development from the original wbccPredictior (Iterson, 2017).

Functions: 1. train_wbcc(data, covariates, cellPercentages, model, ncomp, keep, ...) uses pslr() to train a cell count predictor. 2. predict_wbcc(predictor, data, covariates, transformation, ncomp, impute, ...) predicts cell count values in a new sample. 3. plot_wbcc(measured, predicted) can be used as a validation plot.



ljsinke/wbccPredictor documentation built on May 15, 2019, 12:07 a.m.