SORS: From bulk to spatial metabolomics: bioinformatic methods with applications to glioblastoma

Speaker: Macha Nikolski, CNRS Cellular Genetics Institut, Bordeaux Abstract: Understanding tumor metabolism increasingly depends on the ability to analyze complex omics data rather than on any single experimental modality alone. In glioblastoma, this is particularly true for studying metabolic plasticity across both controlled perturbation experiments and spatially heterogeneous tumor tissues. However, existing bioinformatic methods remain limited and represent a number of particular challenges that we will cover in this presentation as well as the ways to overcome them and move in the direction of functional interpretation of metabolomics datasets. In this talk, Macha presents methodological contributions developed in their team around the issue of interpretability of metabolomics. The first, DIMet, is a framework for differential analysis of labeled metabolomics data that supports isotopologue-level comparisons in multifactorial experimental designs and can be extended to integrative analysis with transcriptomics. The second, SpacePath, is a framework for functional analysis of spatial metabolomics and lipidomics data that addresses annotation ambiguity, partial pathway coverage, and signal quality issues through probabilistic and cheminformatics-based strategies. Together, these methods provide robust computational tools for extracting interpretable metabolic programs from both bulk and spatial metabolomics data, with glioblastoma serving as a motivating application. Further information here: https://www.bsc.es/research-and-devel...