P_null: Build Null Distribution of Probability-Based Distance

In pageRank: Temporal and Multiplex PageRank for Gene Regulatory Network Analysis

Description

Build null model for evaluating the significance of interactions by generating random regulator-target pairs.

Usage

P_null(expmat, net, n = 10000, sep = 5, method = c("difference", "mi"))

Arguments

expmat	(matrix) Gene expression matrix.
net	(data.frame) Network, with "reg" and "target" in column name.
n	(numeric) Number of random pairs.
sep	(numeric) Number of bins for calculating marginal/joint probability.
method	(character) Method for calculating probability-based distance, either PXY-PXPY ("difference") or mutual information ("mi").

Value

(ecdf) ECDF of null distribution.

Author(s)

DING, HONGXU ([e-mail])

Examples

library(bcellViper)
data(bcellViper)
dset <- exprs(dset)
net <- do.call(rbind, lapply(1:10, function(i, regulon){
  data.frame(reg=rep(names(regulon)[i], 10),
             target=names(regulon[[i]][[1]])[1:10],
             direction=rep(1, 10),
             stringsAsFactors = FALSE)}, regulon=regulon))
P_null(dset, net, n=100, method="difference")

pageRank documentation built on Nov. 8, 2020, 6:52 p.m.