Julie Bourbeillon is seeking a fellow for MSCA PF 2026 in bio data mining at L’Institut Agro, A[...]

Description of the project

My research mainly focuses on applying semantic web techniques to develop methods and tools that streamline the acquisition, management, integration, and interpretation of biological data, with a special emphasis on plant biology. I address a range of challenges, such as handling heterogeneous datasets—often mixing qualitative and quantitative variables—optimizing dimensionality reduction, and designing effective ways to visualize complex results. Central to my work is leveraging semantic web techniques, and in particular ontologies, to represent real‑world biological knowledge in order to enhance data consistency and analytical robustness.

Among recent works is the introduction of a novel semantic distance metric that accurately encompasses both quantitative and qualitative variables, enabling a richer depiction of relationships between individuals (e.g., phenotypic descriptions). This semantic distance is integrated into a complete tool that takes biologists from raw data input all the way to intuitive visualization, thereby simplifying the adoption of advanced semantic approaches and enhancing the overall efficiency and clarity of plant biology research.

Another contribution is a pipeline tailored to describe a factor (a qualitative variable of interest) in heterogeneous datasets consisting of qualitative and quantitative variables. This pipeline separately analyses the variables related to the factor using appropriate statistical tests and offers an interactive visualisation that describes the factor.

As part of the MSCA proposal I would like to focus on some of the problems posed by accessing large, heterogeneous plant datasets:

1. Identifying individuals: helping taxonomists resolve ambiguities in the naming of taxa between different reference systems, or even in the absence of a reference system.
2. Consolidating vocabulary: proposing an automatable method for reconciling the terms used in different datasets to describe measurements and modalities of qualitative variables.
3. Integrating data: jointly processing data from different scales (spatial, temporal or related to the experimental context) by matching similar individuals and variables, while taking into account differences in context through appropriate mathematical or statistical transformations.

I hypothesise that the combination of ontologies, repositories, and artificial intelligence methods applied to the underlying problems (ontology alignment, automatic language processing, etc.) can help answer these questions and assist biologists in better understanding datasets.

In our team, we view computer science and data mining as essential tools for assisting biologists in characterizing plants and gaining new insights from their data. We particularly enjoy discovering innovative methods and developing new concepts.

AGRO and other local research institutions organise a Springboard toward the Marie Sklodowska Curie programme. Selected candidate researchers will have the opportunity to attend an online intensive training and coaching course on how to write a successful proposal. This masterclass is a pathway of support for candidates and their supervisors in preparing project proposals, including prescreening activity of draft proposals and the organization of dedicated B2B meetings, by the European Research Office of the University of Angers. More information is available at

Keywords

Computer Science – Semantic Web – Ontologies - Data mining - Data Integration - Dimensionality reduction – Data Visualisation – Plant Biology

Eligibility

Applicants must comply with the mobility rule (having stayed in France for less than 12 months in the past 3 years before the 09th of September 2026). Applicants also must have a maximum of 8 years of full‑time research experience after graduating from their (first) PhD.

We encourage you to apply ASAP; if we receive applications that have a fitting profile, we will close the offer in Euraxess.