Gene Ontology (GO) annotations
======

**Note**: The GO terms are derived from the mapping of UniRef entries to GO
terms, provided in the UniRef data release. See uniref/.

The Gene Ontology (GO) is a knowledgebase of biological functions of genes.

 - Website: http://geneontology.org/

 - Citation: The Gene Ontology resource: enriching a GOld mine. Nucleic Acids
   Res. Jan 2021;49(D1):D325-D334.

GO release 2022-05-16 was used to construct the knowledge graphs.

 - Source: http://release.geneontology.org/2022-05-16/index.html

Inconsistent records between UniRef and GO data releases have been corrected.

Domains:

The GO terms are divided into three domains:

 - function:  molecular function (GO:0003674)
 - process:   biological process (GO:0008150)
 - component: cellular component (GO:0005575)

In addition, there is:

 - all: all three domains combined

Statistics:

 - Number of mapped UniRef entries: 17,214,447
   - component: 7,922,816
   - function: 13,469,279
   - process:   7,833,641

 - Number of unique GO terms: 6,246
   - component:  386
   - function: 3,833
   - process:  2,027

 - Number of unique GO terms, including ancestral ones: 8,750

Database files:

 - uniref/: Mappings of UniRef entries to GO terms, curated to match current
   GO release.

 - go_name.txt: Names (descriptions) of all relevant GO terms (original,
   parents, generic).

 - generic/: Mappings of GO terms to the "generic" subset (a.k.a., GO slim).

   - Note: GO terms can be grouped into a refined set of GO terms, namely GO
     subsets, or GO slims. See official documentation:

     * http://geneontology.org/docs/go-subset-guide/

     Be aware that generic terms may be nested (i.e., one term may be a child
     of another). See below for explanation.

 - parent/: Mappings of GO terms to their parent GO terms. This includes all
   orginal GO terms and their ancestral GO terms.

   - Note: The hierarchical relations among GO terms constitute graphs, instead
     of trees. In this structure, one child term can be mapped to one or more
     parent terms. This structure is a directed acyclic graph (DAG).

     There are no fixed "ranks" in the graph. The three domains (see above) are
     the top-level "roots" of the graphs. See official documentation:

     * http://geneontology.org/docs/ontology-documentation/#the-go-graph

 - idmaps/: Mapping of GO terms to external databases:

   - EC, KEGG, MetaCyc, Reactome, and Rhea

 - curation/: Changes of UniRef-derived GO terms to match current GO release.

Collapsing order:

```
  ORF > UniRef > GO > generic (slim)
                  v
               parent GO > ...
```