#35 · Metabolomics

Tissue-resolved metabolic models, spatial lipidomics, and curated MASH transcriptomics

The common thread is computational infrastructure for metabolic omics: each item converts heterogeneous molecular measurements into a resource for reuse and interrogation. The Drosophila study reconstructs 32 tissue-specific genome-scale metabolic models from single-nuclei transcriptomics and applies muscle-model flux analysis to identify high sugar diet-induced dysregulation at reaction and pathway levels. LipidLocator is an open-source Shiny application that integrates spatial lipidomics MSI normalization, clustering, differential abundance, and MS/MS lipid annotation. MASH-GA curates 45 human and 126 mouse MASH transcriptomic datasets with WGCNA and visualization tools for cross-species comparison. The distinguishing element is the explicit coupling of tissue resolution with pathway-level inference in the Drosophila work, and workflow consolidation in LipidLocator. The items are adjacent rather than mechanistically linked, so the shared element is methodological rather than a common biological finding.