Ontology-based cross-species integration and analysis of Saccharomyces cerevisiae phenotypes

Ontologies are widely used in the biomedical community for annotation and integration of databases. Formal definitions can relate classes from different ontologies and thereby integrate data across different levels of granularity, domains and species. We have applied this methodology to the Ascomycete Phenotype Ontology (APO), enabling the reuse of various orthogonal ontologies and we have converted the phenotype associated data found in the SGD following our proposed patterns. We have integrated the resulting data in the cross-species phenotype network PhenomeNET, and we make both the cross-species integration of yeast phenotypes and a similarity-based comparison of yeast phenotypes across species available in the PhenomeBrowser. Furthermore, we utilize our definitions and the yeast phenotype annotations to suggest novel functional annotations of gene products in yeast.

Background
Yeast phenotypes have been proven useful for investigating and revealing various aspects of cellular physiology and mechanisms. The study of these phenotypes has direct implications for understanding mammalian physiology in the context of pharmacodynamics and pharmacokinetics studies, in understanding signalling and regulatory networks, in studies that focus on the identification of response regulators, activators and inhibitors, and in chemical genetics [1-4]. It is therefore essential that efficient ways are set in place to collect and analyse yeast phenotype data as well as compare them with other organism phenotypes held in a variety of resources.

Over the last years, a plethora of phenotype ontologies has been proposed [5-11]. These ontologies are developed by a variety of biomedical communities and aim to support the annotation of phenotypic observations derived either from the literature or from experimental studies, including large scale phenotype studies [12,13]. To unify the species-specific efforts in representing phenotypes, to enable the integration of phenotype information across species, and to enhance the formally represented genotype-to-phenotype knowledge, a species and domain independent method for decomposing phenotypes was proposed based on the Phenotype And Trait Ontology [14]. This method has been successfully applied both for the direct annotation of species-specific phenotypes and for defining classes in species-specific phenotype ontologies to enable cross-species phenotype integration [15-18].

The Saccharomyces Genome Database (SGD) [19] collects and curates yeast-related phenotype data using the yeast-specific Ascomycete Phenotype Ontology (APO) [20]. Here, we report our efforts to apply the EQ-based method to the APO and enable the reuse of biomedical reference ontologies to describe yeast-related phenotype information as well as integrate it with other species. We apply the results of our analysis to the cross-species phenotype network PhenomeNET [21] and make both the cross-species integration of yeast phenotypes and a similarity-based comparison of yeast phenotypes across species available in the PhenomeBrowser [22].