Recent technological advances in biology now allow for the mass gathering of data, that help describe precisely the main functions of life (epigenome, transcriptome, proteome, metabolome…).
These data contribute to an extensive knowledge of the characteristics of each individual, which is to say its phenotype, the result of the expression and regulation of its genome.
The characterisation of the phenotype of each individual involves storage in a phenotypic database, whose structure and exploitation showed the need for a standardised language capable of defining traits unequivocally. This language would have to be able to be used by many different types of users (geneticists, physiologists, biochemists, producers…)
The construction of this language makes use of ontologies, which are structured collections of terms and concepts concerning a given field, organized to meet the needs of users. Although the ontological approach is already well-advanced in animal biology for several academic species such as the mouse, no ontology is exclusively dedicated to livestock (fishes, poultry, mammals).
Regarding this issue, the PHASE division of INRA, in France, in collaboration with international partners, decided in 2009 to work towards building an authoritative ontology named ATOL (for “Animal Trait Ontology for Livestock”). It should define and organise phenotypical traits of livestock, while taking into account societal concerns and the main types of production (milk, eggs, meat, fertility and nutrition).
Furthermore, when used with semantic search tools, ATOL will allow for more relevant and powerful bibliographical research on livestock phenotypes from a reference corpus from the web. ATOL should also help establish more precise genotype/phenotype relations and associate a phenotype to each allele, while better understanding the variability of physiological plasticity and epigenetic marks, due to the environmental context of the animal. This endeavor should pave the way for an approach of animal selection that is no longer exclusively multi criteria, but rather systemic. From the collection of traits defined by ATOL, through the extraction of information by way of semantic analysis and meta-analysis, one can consider modelling the theoretical working of livestock in an environmental context. This modelling work will allow to optimise the number of relevant traits needed in each genetic selection programme, thanks to a predictive approach.
Aims:
- To have an authoritative ontology on livestock phenotyping, shared by the international scientific and educational community
- To have a language used by computer programs (database management, semantic analysis, modelling…)
- To have the most common traits possible regarding livestock
- To make the ontology be the most operational it can be, and make it close to measure techniques
- To structure the base towards animal production