Section: New Results

Scalable methods to query data heterogenity

Participants : Guillaume Alviset, Olivier Dameron, Xavier Garnier, Vijay Ingalalli, Marine Louarn, Yann Rivault, Anne Siegel, Denis Tagu.

Ontology design and integration [O. Dameron, Y. Rivault] We have contribued to several technics improving data integration in onthologies

• The ATOL ontology [link to ontology] supports the annotation of phenotype traits in livestock. It was extended with health-related traits. For each organism, livestock diseases are organized according to their type (infectious, genetic, metabolic,...), their transmission and their symptoms. [32]

• queryMed is an R package [url] that provides both high-level and low-level functions for facilitating the integration of reference ontologies and datasets represented in RDF as Linked Data. It currently focuses on drugs indications, interactions and contra-indications by integrating the Drug Indication Database (DID) and the Drug Interaction Knowledge Base (DIKB). Typical applications concern public health and pharmaco-epidemiology. [27], [26]

Using AskOmics to integrate heterogeneous data [O. Dameron, A. Siegel]

• We contributed to the conversion of an Alzheimer's disease map into a heavyweight ontology, the Alzheimer's Disease Map Ontology (ADMO, [url]), an ontological upper model based on systems biology terms. It provides the ontological formalization for the existing disease map AlzPathway that gives a detailed and broad account of Alzheimer's Disease pathophysiology [25], [20].

• We also contributed to decipher the role of small non-coding RNAs in the regulation of animal reproduction, especially the role of miR-202 in female fecundity by regulating medaka oncogenesis [16].

Graph compression and analysis [L. Bourneuf]. We introduced a general approach combining procedural and logical languages to specify graph objects. This is a generalization of previous work [37], using the reconstruction of Formal Concept Analysis framework example to target the AI community [23].