shadow: Shadow analysis for msviper objects

In chrischen1/viper_copy: Virtual Inference of Protein-activity by Enriched Regulon analysis

Description

This function performs shadow analysis on msviper objects

Usage

shadow(mobj, regulators = 0.01, targets = 10, shadow = 0.01, per = 1000,
  nullmodel = NULL, minsize = NULL, adaptive.size = NULL,
  iterative = NULL, seed = 1, cores = 1, verbose = TRUE)

Arguments

mobj	msviper object generated by msviper
regulators	This parameter represent different ways to select a subset of regulators for performing the shadow analysis, it can be either a p-value cutoff, the total number of regulons to be used for computing the shadow effect, or a vector of regulator ids to be considered
targets	Integer indicating the minimum number of common targets to compute shadow analysis
shadow	Number indicating the p-value threshold for the shadow effect
per	Integer indicating the number of permutations
nullmodel	Null model in marix format
minsize	Integer indicating the minimum size allowed for the regulons
adaptive.size	Logical, whether the target weight should be considered when computing the regulon size
iterative	Logical, whether a two step analysis with adaptive redundancy estimation should be performed
seed	Integer indicating the seed for the permutations, 0 for disable it
cores	Integer indicating the number of cores to use (only 1 in Windows-based systems)
verbose	Logical, whether progression messages should be printed in the terminal

Value

An updated msviper object with an additional slot (shadow) containing the shadow pairs

See Also

msviper

Examples

data(bcellViper, package="bcellViper")
sig <- rowTtest(dset, "description", c("CB", "CC"), "N")$statistic
dnull <- ttestNull(dset, "description", c("CB", "CC"), "N", per=100) # Only 100 permutations to reduce computation time, but it is recommended to perform at least 1000 permutations
mra <- msviper(sig, regulon, dnull)
mra <- shadow(mra, regulators=10)
summary(mra)

chrischen1/viper_copy documentation built on May 13, 2019, 6:52 p.m.