Centrality-based KEGG enrichment analysis

Usage

retrieve_all_kegg_graphs(organism = "hsa")

ora_kegg_topology(
  genes,
  universe = NULL,
  centrality = igraph::degree,
  organism = "hsa",
  pl = NULL,
  nperm = 1000,
  min_hits = 3,
  min_size = 5,
  max_size = 2500,
  verbose = TRUE
)

gsea_kegg_topology(
  s,
  centrality = igraph::degree,
  null_side = c("both", "right", "left"),
  abs = TRUE,
  organism = "hsa",
  pl = NULL,
  nperm = 1000,
  min_size = 5,
  max_size = 2500,
  verbose = TRUE
)

Arguments

organism: KEGG organism code.
genes: A vector of genes, must be in the EntreZ ID type.
universe: A vector of universe genes. If it is not specified, the total genes in the KEGG pathways are used.
centrality: Centrality method. The value should be a function which accepts an igraph object and returns a vector of centrality values.
pl: A list of KEGG pathways as igraph objects. If the organism is one of "hsa", "mmu" or "rno", the corresponding pl is already generated and will be loaded automatically. For other organisms, use retrieve_all_kegg_graphs() to generate one.
nperm: Number of permutations.
min_hits: Minimal number of the overlapping genes in genes and pathways.
min_size: Minimal number of genes in pathways.
max_size: Maximal number of genes in pathways.
verbose: Whether to print messages?
s: A numeric vector of gene scores. Names must be in the EntreZ ID type.
null_side: How to calculate p-values?
abs: Whether use absolute values of s?

Details

The following lists several useful centrality measures, written as functions. They can be assigned to the centrality argument.

in-degree: function(g) igraph::degree(g, mode = "in")
out-degree: function(g) igraph::degree(g, mode = "out")
betweenness: igraph::betweenness
page rank: igraph::page_rank
in-reach: function(g) CePa::reach(g, mode = "in")
out-reach: function(g) CePa::reach(g, mode = "out")

And many more in the igraph and CePa packages.

ora_kegg_topology() is the ORA-extension.

gsea_kegg_topology() is the GSEA extension. The geneset-level score is calculated as mean(abs(s*w)).