Pan-cancer metabolic landscapes: A multi-omics view

Abstract

Metabolic reprogramming fuels cancer progression, but whether common metabolic patterns exist across tumor types remains elusive. To address this, we developed parseMetab, an R package that integrates large-scale proteomic, transcriptomic, and spatial transcriptomic datasets from 24 human cancers and 3,226 samples to map pan-cancer metabolic dysregulation. Strikingly, glycan biosynthesis pathways - especially those driving fucosylation and sialylation - were consistently upregulated, while early sugar nucleotide precursors were suppressed. Nucleotide metabolism was broadly enhanced, uncovering conserved metabolic programs that enable tumor growth, adaptation, and immune evasion. Our results highlight universal metabolic vulnerabilities that may be therapeutically exploited across diverse cancer contexts. Significance statementWhile metabolic reprogramming is a known hallmark of cancer, identifying universal patterns across diverse tumor types remains a major challenge. We developed parseMetab, a computational framework to integrate large-scale proteomic and transcriptomic data from over 3,200 samples across 24 human cancers. Our analysis reveals a striking, conserved "glyco-switch": tumors consistently prioritize the production of complex, immune-evading surface sugars (fucosylation and sialylation) while depleting their own internal sugar precursor pools. Furthermore, we demonstrate that nucleotide metabolism is broadly hyperactivated across nearly all studied malignancies. By defining these ubiquitous metabolic signatures, this study uncovers conserved vulnerabilities that transcend specific cancer types, providing a map for developing broad-spectrum metabolic therapies to combat tumor growth and immune evasion.