Abstract: Modern molecular methods enable the simultaneous measurement of thousands and thousands of biological states. These newly available genomic, proteomic, and metabolomic data sets are valuable in that they reveal new aspects of biological systems that can be directly measured. But the true potential of this new data to enable fundamental advances in our understanding of human biology and medicine may lie in its use inferring the dynamics of biological systems: how quickly are these biological states changing, and what conditions or interventions control these rates of change? An understanding of biological dynamics has proven essential to disparate areas of medical diagnosis and treatment. For example, experiments revealing white blood cell kinetics guided early success in HIV treatments, and knowledge of hemoglobin glycation rates and blood cell turnover is crucial to current best practices for managing diabetic patients. Kinetics and dynamics cannot be measured directly and must be inferred using computational and mathematical modeling. Because very few clinically-informed investigators have necessary mathematical and computational expertise, most dynamic aspects of human biology and disease remain poorly understood, and patients are unable to benefit from the fundamental diagnostic and prognostic insights this dynamical understanding would enable. I will develop a clinically-informed mathematical and computational framework to infer the dynamics of cellular pathophysiologic processes in humans in vivo using routinely available ensemble measurements of cellular population characteristics. I will apply the modeling framework to all blood cell lineages including lymphocytes, neutrophils, erythrocytes, and platelets and will reveal insights and applications for representative types of disease including cancer (leukemia), infection (sepsis), and autoimmune disease (idiopathic thrombocytopenic purpura). I will synthesize existing scientific and clinical knowledge of cellular pathophysiology into mathematical models describing rates of cellular birth, death, influx, and efflux, as well as how these rates vary among individual patients and within patient cell populations as a function of cell size, age, nuclear complexity, and other single-cell characteristics. I will then compare model parameter trajectories for healthy individuals and patients with disease to reveal new details of disease mechanisms and the pathologic responses they and their treatments elicit. Because the structure of the mathematical models is informed by current knowledge of pathophysiology, model parameters represent personalized quantification of important homeostatic processes and provide new conceptual insights into human pathophysiology. Because models are built with routinely available clinical measurements, these insights will often be immediately translatable. Public Health Relevance: The proposal develops a new mechanism-based modeling framework that will use existing clinical laboratory tests to provide earlier, more accurate, and personalized diagnosis and treatment monitoring for a range of diseases including cancer, infection, and autoimmunity.
Showing the most recent 10 out of 14 publications
