Gene-weighted pathway significance analysis

Description

Test the significance of pathways in microarray experiments. This includes a network-based gene weighting algorithm for pathways. Classical and gene-weighted versions of gene set analysis approaches are both used. When required, this function also corrects for gene weighting biases caused by multiple-subunit protein.

Usage

1
GSE.Test.Main(gExprs.obj, gsets, gNET, check.exprs = TRUE, msp.groups, size.min = 15, size.max = 500, permN = 1000, randN = 30, permFDR.cutoff = 0.5, output.label = "", msp.correction = TRUE)

Arguments

gExprs.obj

Gene expression experiment data object.

gsets

A list of gene sets.

gNET

A gene association network stored in a list.

check.exprs

Logical (TRUE by default). Check and correct the missing values and scaling in the gExprs.obj. If the scale is natural, it will be converted to log2.

msp.groups

A list of multi-subunit proteins.

size.min

Minimum size of gene sets used for analysis. By default 15 genes.

size.max

Maximum size of gene sets used for analysis. By default 500 genes.

permN

Sample permutation times. By default 1000 times.

randN

Gene randomization times. Can be set smaller (say, 30) if you do not care randomization-based significance so as to be faster.

permFDR.cutoff

Sample permutation FDR cutoff. A number between 0 and 1. Set it larger if wish to see the significance of more gene sets.

output.label

A label to name output files, e.g. "P53\_C2".

msp.correction

Logical (TRUE). Whether to do a correction for multi-subunit proteins in gene weighting.

Value

It will write analysis results to .csv files.

Author(s)

Zhaoyuan Fang, Weidong Tian and Hongbin Ji

References

Zhaoyuan Fang, Weidong Tian and Hongbin Ji. A Network-Based Gene Weighting Approach for Pathway Analysis. Submitted.

Examples

1
2
3
4
# Not to run
# library(GANPAdata)
# data("gExprs.p53", "gsets.msigdb.pnas", "gNET", "msp.groups", package="GANPAdata")
# GSE.Test.Main(gExprs.obj=gExprs.p53, gsets=gsets.msigdb.pnas, gNET=gNET, check.exprs=TRUE, msp.groups=msp.groups, size.min=15, size.max=500, permN=1000, randN=30, permFDR.cutoff=0.5, output.label="P53\_C2", msp.correction=TRUE)