W493
Information System for Phenotype Data: Trait, Trials and Genotypes Identification and Integration in GnpIS

Date: Saturday, January 11, 2014
Time: 10:30 AM
Room: Pacific Salon 2
Cyril Pommier , INRA - URGI, Versailles, France
Michael Alaux , INRA - URGI, Versailles, France
Sophie Durand , INRA - URGI, Versailles, France
Raphael Flores , INRA - URGI, Versailles, France
Aminah Keliet , INRA - URGI, Versailles, France
Erik Kimmel , INRA - URGI, Versailles, France
Célia Michotey , INRA - URGI, Versailles, France
Nacer Mohellibi , INRA - URGI, Versailles, France
Dorothee Valdenaire , INRA URGI, Versailles, France
P. Roumet , AGAP - INRA, Montpellier, France
Delphine Steinbach , INRA - URGI, Versailles, France
Hadi Quesneville , INRA - URGI, Versailles, France
GnpIS is an integrative information system dedicated to curated data integration for plant and their bio aggressors. It integrates and links data from different thematic like genomic, expression, genetic resources, polymorphism, phenomic and association studies. The quality of data integration between those thematics relies on a consistent identification of some pivot objects among which the more notables are genetic resources or germplasms, traits and markers.
The phenotype thematic is handled in GnpIS through the Ephesis web application. It allows the integration of multifactorial phenotyping Trial designed for Genotype by environment studies, for either annual or perennial plants and integrates data for wheat, forest trees, grape or fruit trees. This genericity is allowed by the ontology driven design of the information system. Indeed, traits, environment observations and trial levels (bloc, plot, plant, …) are defined by either specific ontologies or by reference ontologies. Each level defines the elements characterized in the trial, like a particular plot or plant. Those elements can receive multiple factors and observations. There is at least one mandatory genotype factor plus as many treatments as needed. The observations made on the trial elements include phenotype or environment observations, and can optionally be dated to set up temporal series and kinetics. The web application allows the set-up of multi trial data sets which can be exported to conduct further analysis either to decipher genotypes to environmental constraints or to prepare association studies.

The possibility to set up such dataset relies on data consistency and therefore GnpIS is based (i)on genotype identification conducted in genebanks like INRA’s BRC and (ii)on ontologies and standard format like those developed in the Transplant project.

Currently, GnpIS integrates many private data plus 15 public trials including 12 Forest Trees trials conducted on 13 years. This covers more than 120 000 elements (plants and plots) and more than 1.4 million phenotypes.

Acknowledgments: GnpIS.Ephesis is built from user feedback provided by agronomist, researcher, and engineers form INRA. It is an official Phenotype data repository for INRA and several national projects. Thanks to those partners help, Ephesis is under continuous improvement to increase usability by the wider possible scientific audience.