PIDC: Inferring Gene Regulatory Network (GRN) using partial...

In hmutpw/PIDC: Infering Gene Regulatory Network (GRN) using partial information decomposition and context (PIDC) method

Description

The PIDC package implements the method in paper (2017 Chan et al.) with few modifications. Besides, The code used for calculating the Multivariate Information (MI) and specific.information, discretizing gene expression data were partly adopted from package Informeasure.

Usage

PIDC(
  expMat,
  regulators = NULL,
  targets = NULL,
  logScale = FALSE,
  ncores = 1,
  diag = c("auto", "zero", "one"),
  verbose = interactive()
)

Arguments

expMat	A matrix storing the gene expression data. Rows corresponding to features (eg. gene symbols) and columns corresponding to samples (cells). Raw read counts, UMIs, TPMs or logNormalized counts were supported.
regulators	The regulator genes used for GRN inferring (eg, transcription factors). At least two regulators required. Default: NULL, using all the genes in gene expression matrix. Default: NULL, all genes.
targets	The target genes used for GRN inferring. Default: NULL, using all the genes in gene expression matrix. Default: NULL, all genes.
logScale	Whether to log-scale the input data? Default: FALSE.
ncores	Number of cores used for parallel calculation. The running time heavily depend on the number of regulators and targets, for genes over 5000, we strongly suggest you to use multi-cores. Default: 1.
diag	Numeric. The weight in the diagonal of output square matrix. Default: 1.
verbose	Whether to print message while running. Default: TRUE.

Value

A matrix with weighted value between regulators and targets.

Examples

data(expMat)
PIDC_res <- PIDC(expMat)
head(PIDC_res)

hmutpw/PIDC documentation built on Jan. 8, 2022, 12:20 a.m.