Mapping and computation of dynamic competition, in life and disease

Genome-wide metabolic maps have facilitated the design of chemical production methods and disease analysis but are limited by lacking realistic constraints like V_max for enzymes and non-linear effects such as cooperation and competition. This thesis extends metabolic maps to include competition between metabolites and cells, enhancing the understanding of diseases like phenylketonuria (PKU) and cancer, and aiding in the discovery of nutritional and other drug therapy targets.
Chapter 2 focuses on multicellular system maps and dynamic competition Flux Balance Analysis (dcFBA). It shows that cell competition can lead to outcompeting others unless regulation is enforced, suggesting social interactions are crucial for multicellularity. In cancer, it’s the loss of social traits, not faster growth rates, that allows cells to dominate, indicating that therapies enhancing cross-regulation may be more effective than targeting growth rates. Chapter 3 introduces a competition-based FBA model to explain PKU, revealing how phenylalanine’s interference with brain function leads to pathology, and why phenylalanine reduction is more effective than tyrosine supplementation in treatment.
Chapter 4 uses transcriptome sequencing data to constrain metabolic maps for hepatocellular carcinoma and normal liver cells, uncovering differences in nutrient dependency between cancer and healthy cells. Chapter 5 applies single-cell transcriptomics to build V_max-constrainted and identify metabolic vulnerabilities in liver cancer, predicting potential drug targets for therapeutic intervention.
This thesis explores multiple extended uses of metabolic maps. The incorporation of competition into metabolic maps enhances our ability to interpret disease and design treatments.