epicMI

Using the unreliability_MI function, calculate normalized mean intensities (MI), estimate unreliability and scores and identify threshold for removing Infinium probes on the Illumina MethylationEPIC microarray v1.0.

Requirements

R >= 3.5.0
minfi >= 1.42.0
stringr
dplyr
ggpubr
ggplot2

Installation

if(!require(devtools)) install.packages("devtools")
devtools::install_github("ChVav/epicMI")

Example

Loading package and preparing RGset:

library(epicMI)
RGset <- read.metharray.exp(targets)

out <- unreliability_MI(RGset, samples, grid_max_intenisty = 5000, grid_step = 100, number_beta_generated = 1000)

By default, unreliability_MI function calculate normalized mean intensities (MI), estimate unreliability and scores and identify threshold for removing probes estimate unreliability and calculate MI scores on all samples, using p-noise method, with estimation on Reliability Map grid: M(0,'grid_max_intenisty') x U(0,'grid_max_intenisty') (where by default 'grid_max_intenisty' = 5000), with 'grid_step' = 100 and 'number_beta_generated' = 1000:

out <- unreliability_MI(RGset, samples)

Note: Method can be apply to other Illumina Methylation microarrays (450k, EPICv2.0 etc)