| uniprot_background {GCModeller} | R Documentation |
uniprot_background builds a GSEA background model by integrating: \itemize{ \item A protein table exported from UniProt (e.g. via the “proteinTable” tool). \item KEGG gene–KO mappings obtained from the KEGG REST API link/ko endpoint. } The function maps proteins (rows in proteinTable) to KEGG Orthology (KO) identifiers, then to KEGG reference pathways, and finally constructs a background object suitable for enrichment analysis with the gseakit framework.
uniprot_background(proteinTable, ko.maps,
id.key = 'row.names',
species.code = 'map',
cluster.idset = NULL);The function performs the following steps: \enumerate{ \item If ko_maps is a directory, reads all .txt files inside it, expecting them to be KEGG link/ko results (tab-delimited: kegg_id \t KO), and combines them into a single data frame. \item Loads a standard reference KO set via background::KO_reference() and converts it to a gene set list using as.geneSet. \item Groups the kegg_id–KO mapping by KO and creates a lookup list from each KO to its associated KEGG gene IDs. \item Normalizes proteinTable into a data frame and ensures the identifier column specified by id_key exists (using row names if requested). \item For each KEGG reference pathway (cluster): \itemize{ \item Parses the pathway ID and name from the cluster identifier (expected format "pathway_id - pathway_name"). \item Retrieves the set of KO identifiers associated with that pathway. \item Uses these KO identifiers to look up the corresponding KEGG gene IDs from the ko_maps lookup. \item Maps those KEGG gene IDs to rows in proteinTable and builds a data frame with columns: xref (gene identifier), name, alias, KEGG, uniprot. \item Wraps this data frame as a gsea_cluster object. } \item Collects all non-NULL clusters and converts them into a Background object via as.background. }
An object of class Background from the gseakit framework, which represents a GSEA background model (see as.background). Conceptually, the returned object is a collection of clusters, where each cluster corresponds to a KEGG pathway or map, and each cluster entry contains protein-level annotations (identifier, name, alias, KEGG, UniProt) for the genes in that pathway.
#\dontrun{
# Example: build background from a UniProt protein table and a directory of
# KEGG link/ko results
# 1) UniProt protein table (exported via the UniProt "proteinTable" tool)
prot_tbl <- read.delim("uniprot_proteins.tsv")
# 2) Directory containing files from link_ko or direct KEGG REST calls, e.g.
# https://rest.kegg.jp/link/ko/taes:803091+taes:803092+taes:123456
ko_dir <- "./kegg_ko_links"
# 3) Build background model
bg <- uniprot_background(
proteinTable = prot_tbl,
ko_maps = ko_dir,
id_key = "row.names" # use row names of prot_tbl as identifiers
)
# Inspect the background object
str(bg)
#}