What we can learn about Escherichia coli through application of Gene Ontology

Abstract

How we classify the genes, products, and complexes that are present or absent in genomes, transcriptomes, proteomes, and other datasets helps us place biological objects into subsystems with common functions, see how molecular functions are used to implement biological processes, and compare the biology of different species and strains. Gene Ontology (GO) is one of the most successful systems for classifying biological function. Although GO is widely used for eukaryotic genomics, it has not yet been widely used for bacterial systems. The potential applications of GO are currently limited by the need to improve the annotation of bacterial genomes with GO and to improve how prokaryotic biology is represented in the ontology. In this review, we will discuss why GO should be adopted by microbiologists, and describe recent efforts to build and maintain high-quality GO annotation for Escherichia coli as a model system.

The EcoCyc Database

“…GO and its applications are described in more detail in [56]. Since the summer of 2008, we have been periodically generating a file containing all E. coli K-12 GO term annotations, called gene_association.ecocyc, that may be obtained from the Gene Ontology Consortium.…”

EcoGene 3.0

“…All the GO terms currently associated with E. coli K-12 genes and the GO hierarchy of terms, names and definitions, including all ancestor and children relations, were obtained from the Amigo website (25). The E. coli GO term assignments at Amigo are compiled from several sources including EcoCyc (26), EcoliWiki (27), UniProtKB (11,28) and InterPro (29). We made two observations consistent with problems reported elsewhere (30).…”

An ontological analysis of some biological ontologies

“…Inadequacies could also be found in the use of relations such as is_A and part_Of (Smith et al, 2005; Burek et al, 2006). The distinction between function and their functioning in the GO has also been confusing, though a solution was attempted by the GO by appending the term “activity,” e.g., “galactokinase activity” (Krummenacker et al, 2009). Another problem which could arise from the use of multiple inheritances and is_A overloading is polysemy (Guarino, 1999).…”

“…The problem of multiple inheritance in its conceptual hierarchies prevents it from logical reasoning applications. To understand one such inadequacy, let's take an example from the GO described graphically in Krummenacker et al (2009). If galactokinase activity is made a subclass of carbohydrate kinase activity and phosphotransferase activity, then as per the rules of subsumption (Guarino, 1998), it would inherit the identity of both the super-classes.…”