Project deliverable Open Access
Tardieu, Francois; Neveu, Pascal; Pommier, Cyril; Pieruschka, Roland; Saint Cast, Clément; Draye, Xavier; Cabrera Bosquet, Llorenç; Usadel, Björn
Phenomic datasets are at the crossroad of several disciplines, in particular plant physiology, genetics, ecology and crop modelling. All these communities have developed their own solutions to organise and store resulting datasets in dedicated repositories or information systems, so it is crucial that the effort of EMPHASIS to organize phenomic datasets results in information systems that are interoperable with those of other infrastructures such as ELIXIR (genomics and genetics), ANaEE (Ecology) and AgMIP (crop modelling). This is in a context in which data sciences are receiving an increasing interest, in particular with the EOSC initiative of the European Commission.
The information systems developed by different communities for phenotypic data differ in their organization and content. Phenomic information systems need to track all elements for re-analysis of individual experiments, in particular the time courses and spatial variability of environmental variables, the time course of traits and the link between samples, organs, genotypes, plants,events, x-y positions of each plant or plot. This requires the use of semantic web to automatically generate complex metadata from a few indications. This level of detail is not necessary in Genetic-Genomic or in Ecology, so corresponding information systems relate integrative plant traits with integrative indicators of environmental conditions, and/or with genomic information. The crop modelling community developed its own standards for traits, environmental variables and managing practices. EMPHASIS has launched common working groups with each of these communities, resulting in a clarification of the role of each information system, in common tasks for mapping ontologies and in a list of tasks that will facilitate interoperability between information systems.