cepa.ora.all: Apply centrality-extented ORA on a list of pathways

Description Usage Arguments Details Value Author(s) Examples

View source: R/ora_extension.R

Description

Apply centrality-extented ORA on a list of pathways

Usage

1
2
3
cepa.ora.all(dif, pc, bk = NULL, cen = default.centralities,
    cen.name = sapply(cen, function(x) ifelse(mode(x) == "name", deparse(x), x)),
    iter = 1000)

Arguments

dif

differential gene list

pc

a pathway.catalogue class object

bk

background gene list. If background gene list are not specified, use whole human genes

cen

centrality measuments, it can ce a string, or a function

cen.name

centrality measurement names. By default it is parsed from cen argument

iter

number of simulations

Details

The traditional over-representation analysis (ORA) to find significant pathways uses a 2x2 contingency table to test the independency of genes belonging to a functional category and these genes being differentially expressed, usually by Fisher's exact test. The ORA only consider the number of genes and the function extend traditional ORA with network centralities.

The differential gene list and the background gene list should be indicated with the same identifiers (e.g. gene symbol or refseq ID). All genes in the differential gene list should exist in the background gene list. If users use the PID.db data, all genes should be formatted in gene symbol.

If the centrality measurement is set as a string, only pre-defined "equal.weight", "in.degree", "out.degree", "degree", "betweenness", "in.reach", "out.reach", "reach", "in.spread", "out.spread" and "spread" are allowed. More centrality measurements can be used by setting it as a function (such as closeness, cluster coefficient). In the function, we recommand users choose at least two centrality measurements. The default centralities are "equal.weight", "in.degree", "out.degree", "betweenness", "in.reach" and "out.reach".

However, in most circumstance, the function is called by cepa.all.

Value

A cepa.all class object

Author(s)

Zuguang Gu <z.gu@dkfz.de>

Examples

1
2
3
4
5
6
7
8
## Not run: 
data(PID.db)
# ORA extension
data(gene.list)
# will spend about 20 min
res.ora = cepa.ora.all(dif = gene.list$dif, bk = gene.list$bk, pc = PID.db$NCI)

## End(Not run)

jokergoo/CePa documentation built on July 2, 2021, 10:28 p.m.