Background Fatty−acid metabolism (FAM) is rewired in bladder cancer (BLCA), yet its impact on intratumoral diversity and patient outcome is unclear. Methods To characterize FAM heterogeneity, we integrated spatial and single-cell transcriptomic approaches. We employed high-dimensional weighted correlation network analysis (hdWGCNA) alongside five distinct enrichment methods (ssGSEA, AddModuleScore, AUCell, singscore, and UCell) to identify modules with elevated FAM activity. Subsequently, machine learning algorithms were applied to bulk RNA sequencing datasets to pinpoint the key gene with highest predictive value. This candidate underwent validation through functional experiments and analysis of clinical specimens. Results Malignant epithelial cells displayed the strongest FAM activity. Cross−platform scoring and co−expression analysis produced a refined high−FAM gene set. Integrating this signature with bulk datasets singled out PRDX1 as a key driver. PRDX1 was up−regulated in tumors, predicted poorer prognosis, and was enriched in malignant epithelial cells. Silencing PRDX1 curtailed BLCA cell proliferation, migration, and invasion. Conclusions PRDX1 emerges as a FAM−linked oncogenic biomarker that fosters BLCA progression. These findings define the metabolic hierarchy of BLCA and nominate PRDX1 as a candidate target for personalized therapy.
Wang et al. (Thu,) studied this question.