Method: Improvement of the animal trait ontology for livestock (ATOL): contribution of European infrastructure projects

Technological advances enable increasingly accurate phenotyping of animals and produce large amounts of data. Using computers to analyse these data and thus compare phenotypes requires a standard language that unambiguously defines phenotypic traits to serve as a worldwide reference for potential users (e.g., geneticists, physiologists, biochemists, modelers, livestock farmers). The absence of such a language/reference for livestock species prompted INRAE, in collaboration with its international partners, to develop an ontology called Animal Trait Ontology for Livestock ( ATOL ). Its aims were to list and organise phenotypic characteristics of livestock species and then classify them into types based on performance traits (e.g., feed efficiency, fertility), production traits (e.g., milk, meat, eggs, fatty liver), and societal traits (e.g., animal welfare). Here, we describe the foundations of ATOL from its inception to the present, highlighting the computer tools that enable the entire scientific community to access and use it and emphasising the crucial role of the European Union projects AQUAEXCEL, SmartCow, PIGWEB, and Intaqt. These projects have expanded ATOL, increasing the number of traits from 1 746 to over 6 000. The future of ATOL looks promising because new tools such as artificial intelligence have the potential to replace manual experimental data annotation. ATOL, which is publicly available online, is expected to be shared worldwide for the common use of unique terms to annotate publications and databases or mine literature, thus promoting both systemic and predictive biology.