Tumor classification and survival estimation using metabolic activity scores from RNA sequencing data

• Activity Number: 384 - Next-Generation Sequencing and High-Dimensional Data
• Type: Contributed
• Date/Time: Thursday, August 12, 2021 : 12:00 PM to 1:50 PM
• Sponsor: Biometrics Section
• Author(s): Jack Goodman* and Marcus Alexander
• Companies: Frank H. Netter SOM and Yale University
• Keywords: cancer; survival; RNA; metabolism; transcriptional pathways; prognosis

Abstract

Recent research on metabolic changes accompanying oncogenesis offer possibilities for improved tumor grading and prognosis based on molecular profiling. Cancer expression data from 5,980 donors across 23 tissue types were collected from The Cancer Genome Atlas (TCGA). Genes were renormalized and metabolic reaction activity scores (RAS) were estimated using relative levels of enzyme subunits and isoforms. K-means clustering was used to classify tumors based on RAS, followed by Cox proportional hazard estimators of patient survival probabilities. Across all donors, longer survival was associated with downregulation of glycolysis and the pentose phosphate pathway—even after controlling for possible confounders. Prolonged survival was also associated with upregulation of fatty acid synthesis and mitochondrial bicarbonate production. Subgroup analysis by TCGA project revealed that, generally, differential regulation of these pathways was conserved. Our findings serve to inform development of molecular biology-informed tumor classifications and aid in metabolism-based drug discovery.