Human-Specific Heterogeneous Network

The tKOI knowledge graph: an undirected igraph object with 939,059 nodes and 10,622,200 edges linking human genes to the concepts they relate to. It is shipped as a plain igraph object and loads lazily the first time tkoi::tkoi_net is used (this takes a few seconds and about 0.5 GB of memory).

Usage

tkoi_net

Format

An igraph object. Vertex attributes:

name: Unique node ID, e.g. "4:c77f6410-...:16050".
identifier: Source identifier (Entrez ID, GO ID, UBERON ID, ...).
source: Source database.
labels: Node type in Neo4j label form, e.g. "['Gene']".
degree: Node degree in the full network.

The edge attribute edge_type names the relationship (e.g. "PARTICIPATES_GpBP"). The node types included are:

Anatomy: Nodes representing anatomical structures and systems.
BiologicalProcess: Nodes for functional biological processes, such as signaling pathways.
CellType: Nodes describing different cell types.
CellularComponent: Nodes for subcellular structures, organelles, and macromolecular complexes.
ClinicalLab: Nodes representing clinical measurements and diagnostic data.
Complex: Nodes for molecular and protein complexes.
Compound: Nodes for chemical compounds, identified by InChIKey, ChEBI ID, or ChEMBL ID (see compound_annotation). run_tkoi reports only those in human_metabolites.
Disease: Nodes for diseases and pathological conditions.
EC: Nodes categorized by Enzyme Commission numbers.
Gene: Nodes for genetic elements, such as genes and genetic markers.
MiRNA: Nodes for microRNAs and their regulatory roles.
MolecularFunction: Nodes describing molecular activities performed by gene products.
Pathway: Nodes representing sequences of molecular interactions and reactions.
Protein: Nodes for protein molecules.
ProteinDomain: Nodes for specific structural or functional domains within proteins.
ProteinFamily: Nodes for groups of evolutionarily related proteins.
PwGroup: Nodes for pathway groups aggregating multiple related pathways.
Reaction: Nodes for biochemical reactions and their participants.

Details

This heterogeneous network integrates multiple biological datasets to represent complex relationships within the human system. It serves as the foundation for network-based analyses in the tkoi package, such as personalized PageRank calculations and enrichment analyses.

Examples

# \donttest{
igraph::vcount(tkoi::tkoi_net)
#> [1] 939059
table(igraph::V(tkoi::tkoi_net)$labels)
#> 
#>           ['Anatomy'] ['BiologicalProcess']          ['CellType'] 
#>                 13770                 12996                  2744 
#> ['CellularComponent']       ['ClinicalLab']           ['Complex'] 
#>                  1708                 59296                  3318 
#>          ['Compound']           ['Disease']                ['EC'] 
#>                554526                 11448                  8764 
#>              ['Gene']             ['MiRNA'] ['MolecularFunction'] 
#>                 19503                  2656                  3569 
#>           ['Pathway']           ['Protein']     ['ProteinDomain'] 
#>                  4831                194076                 14193 
#>     ['ProteinFamily']           ['PwGroup']          ['Reaction'] 
#>                   659                  6343                 24659 
head(igraph::V(tkoi::tkoi_net)$identifier)
#> [1] "UBERON:0003233" "UBERON:2001901" "UBERON:0004321" "UBERON:0002414"
#> [5] "UBERON:2005118" "UBERON:0034769"
# }