Characterizing Cellular Metabolism via Flux Enrichment Scores: A Versatile Framework for Bulk, Single-Cell, and Spatial Transcriptomics

Metabolic reprogramming is a fundamental feature of diverse biological processes, serving as a key indicator of cellular state and adaptation. To capture these dynamic shifts across various scales, we present a versatile computational framework that integrates transcriptomic profiles spanning bulk, single-cell, and spatially resolved datasets into curated metabolic networks. By leveraging Reaction Activity Scores (RAS) and flux sampling, our method generates relative Flux Enrichment Scores (FES). These scores enable the functional characterization and robust comparison of metabolic activity across different biological conditions, cell types, or tissue regions. A central component of this work is the application of spatial Flux Balance Analysis (spFBA), which uncovers region-specific metabolic rewiring and niche heterogeneity within tissue architectures. We demonstrate its ability to identify localized metabolic programs and nutrient exchange patterns that remain invisible to gene expression analysis alone. Furthermore, we showcase the framework’s flexibility across data scales and its potential for integration into mechanistically informed machine learning pipelines to identify non-metabolic transcriptional predictors of metabolic activity. This multi-scale framework provides a comprehensive and accessible toolset for investigating metabolic dynamics in any biological context, facilitating a deeper understanding of cellular identity and metabolic plasticity.