The Bioinformatics platform supports the Unit's research project through two approaches: an artificial intelligence approach and a multi-omics analysis approach.
The multi-omics approach aims to integrate and interpret RNAseq, single cell RNAseq, methylation (Array, RRBS, RRHP), GWAS (genome-wide association study), proeomics (flow cytometry and serological data) clinical data. The analysed data are generated within the laboratory or come from public resources and European collaborative projects (e.g. the H2020-IMI PRECISESADS and 3TR projects). The tools used are based on the R language and various packages. The unit works in collaboration with the ABIMs platform in Roscoff, which provides work spaces and various analysis tools.
The artificial intelligence approach covers various fields: from natural language processing for the automation of systematic literature reviews to image analysis for the segmentation and annotation of tissue sections, and data mining to build predictive models of the evolution of pathologies. The unit has two computing machines dedicated to training these algorithms. Several outputs are available on the LBAI github page.
Fig: Molecular pattern distribution is represented by 4 clusters of pSS patients with different canonical pathways
Extract from A new molecular classification to drive precision treatment strategies in primary Sjögren’s syndrome. Perrine Soret, Christelle Le Dantec, Emiko Desvauxet al. June 2021, Volume12 - Nature Communications.