fig.FDR_FNR.boxplot: Boxplot of FDR and FNR for Simulated data

In LXQin/PRECISION.seq: Comparison of Depth Normalization Methods

Description

Function for the boxplots of FDR and FNR based on golden truth (typically the simulated benchmark data set simulated.data()) and the differential expression analysis (typically the results from the simulated normalized test data set simulated.data()).

Usage

fig.FDR_FNR.boxplot(DEA.list, truth.DEA.list, title, subset = NULL)

Arguments

DEA.list	A list with each element as a normalization method including the differential expression analysis (DEA) results from multiple simulated data
truth.DEA.list	A list of DEA results of benchmark data for multiple simulated dataset
title	The title of the figure
subset	optional A vector of a subset of genes/markers for this analysis

Examples

simulated <- simulated.data(proportion = c(0.15, 0.25),  median = c(2, 4), numsets = 10)
norm.methods <- c("norm.TMM", "norm.TC", "norm.UQ", "norm.med")
test.DEA.list <- list()
for (i in norm.methods){
temp = lapply(1:10, function(x) pip.norm.DE(raw = simulated[[x]]$simulated_test,
                                            groups = simulated[[x]]$simulated_group, norm.method = i))
test.DEA.list <- append(test.DEA.list, list(temp))
}
names(test.DEA.list) = norm.methods
benchmark.DEA.list <- lapply(1:10, function(x) DE.voom(RC = simulated[[x]]$simulated_benchmark,
                                                       groups = simulated[[x]]$simulated_group, P = 0.01))
fig.FDR_FNR.boxplot(DEA.list = test.DEA.list, truth.DEA.list = benchmark.DEA.list, title = "Example of FDR and FNR boxplot")

LXQin/PRECISION.seq documentation built on Dec. 18, 2021, 3:41 a.m.