Multiple Facets of Protein Function
The phrase “protein function” is very ambiguous, as the functional role of a protein can be described in many different contexts. It can be described in terms of: (1) the molecular function of the protein; (2) its role in biological pathway(s); or (3) its cellular location. Natural language annotations in databases and the literature are too vague and unspecific to accurately describe the function(s) of a protein. This has led to the development of a common organized protein annotation vocabulary such as the Enzyme Commission (EC) number and Gene Ontology (GO) , which are the most commonly used protein function annotation resources.
The Enzyme Commission (EC) number  system is a numerical classification system for enzymes that uses a hierarchical set of four numbers separated by periods to represent the catalytic reaction that it carries out. For example, the EC number 18.104.22.168 describes an isomerase (EC 5.-.-.-) that acts as an intramolecular oxidoreductase (EC 5.3.-.-) and interconverts aldoses and ketoses (EC 5.3.1.-) using glucose-6-phosphate as the substrate (EC 22.214.171.124).
The Gene Ontology (GO)  is the most comprehensive and widely used resource of protein annotations. GO annotation can be used to assign functional terms to both enzymes and non-enzymes from three structured, non-overlapping ontologies in a species-independent manner: (1) molecular function ontology (MFO) describes the biochemical activity of the protein at the molecular level; (2) biological process ontology (BPO) describes the cellular processes and pathways in which the protein is involved; and (3) cellular component ontology (CCO) describes the compartment(s) of the cell in which the protein performs its action.
Figure 2.1 Function annotations for the mouse protein, glucose-6-phosphate isomerase (Uniprot Accession no. P06745) from the Enzyme Commision (EC) number system and Gene Ontology.
The sources of these annotations can be literature references, experimental results, author statements, database references, or computational outputs.
Figure 2.1 shows the functional annotations for the enzyme Phosphoglucose isomerase (PGI) in the mouse, which can moonlight as a tumor-secreted cytokine and angiogenesis factor and also as a neurotrophic factor. The EC number can only describe the catalytic function of PGI; however, the GO annotations from the three ontologies are sufficient to completely describe the functions of the moonlighting protein (see Figure 2.1).