Skip to contents

GO, KEGG, and Reactome over-representation analysis (ORA)

gly_enrich_go.Rd

Perform GO, KEGG, and Reactome ORA for protein UniProt accessions. For glycoproteomics experiments, the function extracts unique proteins from the variable information.

Usage

gly_enrich_go(exp, add_info = TRUE, ...)

gly_enrich_go_(proteins, ...)

gly_enrich_kegg(exp, add_info = TRUE, ...)

gly_enrich_kegg_(proteins, ...)

gly_enrich_reactome(exp, add_info = TRUE, ...)

gly_enrich_reactome_(proteins, ...)

Arguments

exp

(Only for gly_enrich_go(), gly_enrich_kegg(), gly_enrich_reactome()) A glyexp::experiment() object.

add_info

A logical value. This parameter is included for API consistency but has no effect since enrichment results do not contain variable or sample columns. Only applicable to top-level APIs.

...

Additional arguments passed to clusterProfiler::enrichGO(), clusterProfiler::enrichKEGG(), or ReactomePA::enrichPathway(). See the "Additional arguments" section for more information.

proteins

(Only for gly_enrich_go_(), gly_enrich_kegg_(), gly_enrich_reactome_()) A character vector of UniProt accession IDs.

Value

A list with two elements:

  • tidy_result: A tibble with enrichment results containing the following columns:

    • id: Term ID (GO:XXXXXXX, hsa:XXXXX, or R-HSA-XXXXX)

    • description: Term description

    • gene_ratio: Ratio of genes in the term to total genes in the input

    • bg_ratio: Ratio of genes in the term to total genes in the background

    • p_val: Raw p-value from hypergeometric test

    • p_adj: Adjusted p-value

    • q_value: Q-value (FDR)

    • gene_id: Gene IDs in the term (separated by "/")

    • count: Number of genes in the term

  • raw_result: The raw clusterProfiler enrichResult object The list has classes glystats_go_ora_res/glystats_kegg_ora_res/glystats_reactome_ora_res and glystats_res.

Required packages

These functions require the following packages to be installed:

  • clusterProfiler for enrichment analysis

  • ReactomePA for Reactome pathway analysis

  • org.Hs.eg.db for human gene annotation (GO analysis only)

Details

These functions perform over-representation analysis using the specified database.

gly_enrich_go(), gly_enrich_kegg(), and gly_enrich_reactome() are the top-level APIs that work with glyexp::experiment() objects and extract protein information automatically from the "protein" column in the variable information tibble.

gly_enrich_go_(), gly_enrich_kegg_(), and gly_enrich_reactome_() are the underlying APIs that work with protein vectors directly, providing more flexibility for users who don't use the glyexp package.

Gene Extraction (top-level APIs only): Proteins are extracted from the experiment's variable information. The function looks for columns containing protein identifiers and uses them for enrichment analysis. Protein identifiers should be UniProt accessions.

GO Analysis: Uses clusterProfiler::enrichGO() with UniProt IDs as input.

KEGG Analysis: Uses clusterProfiler::enrichKEGG() with UniProt IDs as input.

Reactome Analysis: Converts UniProt IDs to Entrez IDs and uses ReactomePA::enrichPathway().

Additional arguments

universe can be passed as a character vector of uniprot accession, or a glyexp::experiment() object. If latter, the background proteins will be extracted from the experiment. This can be convenient when you first perform differential analysis and then perform enrichment on the differentially expressed proteins:

dea_res <- gly_limma(exp)
sig_exp <- filter_sig_vars(exp, dea_res)
enrich_res <- gly_enrich_go(sig_exp, universe = exp)

For gly_enrich_reactome(), an OrgDb can be passed through ... to clusterProfiler::bitr() to convert UniProt to Entrez IDs. By default, org.Hs.eg.db is used. This is useful when you want to use a different organism than human:

enrich_res <- gly_enrich_reactome(exp, OrgDb = "org.Mm.eg.db")

See also

clusterProfiler::enrichGO(), clusterProfiler::enrichKEGG(), ReactomePA::enrichPathway()