Poster Presentations
Day 2, June 11（Thu.） Room P (5F 501+502)
- 2P-35
Repository-scale exploration of metabolite changes by structuring differential profiles from public metabolomics data
(1Kyutech, 2RIKEN, 3Keio Univ., 4HMT, 5Kitasato Univ.)
oEisuke Hayakawa1,2, Yuto Koga1, Mikiko Takahashi2, Kozo Nishida2, Rira Matsuta3,4, Hiroyuki Yamamoto4, Shin Kawano5
Mass spectrometry–based metabolomics has generated vast amounts of experimental data across diverse biological systems. Public repositories such as MetaboLights and the Metabolomics Workbench contain thousands of studies covering a wide range of organisms, experimental conditions, and disease states. These datasets represent an exceptionally valuable resource because they capture metabolic responses to numerous biological phenomena. However, repository data remain underutilized, as metabolomics datasets are typically distributed as study-specific intensity matrices with heterogeneous metadata, making systematic comparison across studies difficult.
In this study, we developed a framework to reorganize repository-scale metabolomics data based on differential metabolite profiles. Group-wise comparisons were extracted from public datasets, and metabolic changes were represented using fold change values together with statistical significance.
In this representation, each profile corresponds to a biological contrast, while metabolites constitute the variables describing metabolic responses. By structuring repository data in this way, metabolomics studies can be represented as a large collection of metabolite changes associated with diverse biological phenomena.
This framework enables repository-wide exploration of metabolic responses and provides a foundation for cross-study analysis and data-driven hypothesis generation in metabolomics.