Dissertation
Determining metabolic fluxes from metabolomics and stable isotope tracing data
University of Iowa
Doctor of Philosophy (PhD), University of Iowa
Spring 2024
DOI: 10.25820/etd.007323
Abstract
Metabolic fluxes are the reaction rates of biochemical reactions in a metabolic pathway. This thesis examines the capacity to reliably compute estimates of metabolic fluxes from isotopic non-stationary systems in metabolic steady state. Determining the dynamics of these fluxes is critical to understanding the pathophysiology of diseases associated with disruption of normal metabolism.
We develop a procedure for translating diagrams of metabolic pathways into mathematical models composed of hybrid systems of differential equations and algebraic constraints. This process begins with the use of graph theory to represent biochemical networks as weighted directed graphs. The method applies to linear, branched, and cyclic pathways. We also establish essential models for extending this approach to include changing metabolite concentration as well as isotope dynamics.
Kinetic flux profiling (KFP) is an experimental method for computing fluxes in pathways incorporating isotopic label while in metabolic steady state. The dynamical systems model of this pathway involves parameters describing metabolic fluxes. By re-scaling the dynamical systems model we discover a connection between the diagram of the pathway and the number of parameters required in the model. We use Bayesian parameter estimation on simulated data sets to explore the ability of the method to quantify the model’s parameter values. Bayesian methods allow us to ascertain both the accuracy and the certainty of those estimates. The obstacles preventing precise parameter estimations are articulated using local sensitivity analysis and singular perturbation theory. Parameters associated with metabolites downstream in the pathway are more challenging to estimate.
Finally, we extend the KFP models to include reversible reactions between metabolites. The re-scaled models for reversible reactions share the connection between the number of required parameters and components in the diagram. The difficulties of parameter estimation increase with the need to compute additional fluxes. Again using local sensitivity analysis and singular perturbation analysis, we describe the limitations impeding accurate parameter estimations using the Bayesian method. The investigation of both reversible and non-reversible KFP models revealed that accurate parameter estimations are infeasible when a large separation of timescales exists between the turnover rates of upstream metabolites and their downstream products.
Details
- Title: Subtitle
- Determining metabolic fluxes from metabolomics and stable isotope tracing data
- Creators
- Breanna Guppy
- Contributors
- Colleen Mitchell (Advisor)Eric Taylor (Committee Member)Bruce Ayati (Committee Member)Chad Grueter (Committee Member)
- Resource Type
- Dissertation
- Degree Awarded
- Doctor of Philosophy (PhD), University of Iowa
- Degree in
- Applied Mathematical and Computational Sciences
- Date degree season
- Spring 2024
- DOI
- 10.25820/etd.007323
- Publisher
- University of Iowa
- Number of pages
- xv, 121 pages
- Copyright
- Copyright 2024 Breanna Guppy
- Language
- English
- Date submitted
- 04/23/2024
- Description illustrations
- illustrations, graphs
- Description bibliographic
- Includes bibliographical references (pages 118-121).
- Public Abstract (ETD)
- Metabolic fluxes are the reaction rates of life-sustaining chemical reactions within a cell and metabolites are the components of these reactions. Determining the changes in these fluxes is crucial to understanding chronic diseases with metabolic repercussions. Kinetic flux profiling (KFP) is an experimental method for computing fluxes by introducing isotope-labeled nutrient into a metabolic pathway. Measurements of proportion labeled are taken at multiple time points for each metabolite in the pathway. In this thesis, we simulate data from KFP experiments and use Bayesian parameter estimation to judge the accuracy and reliability of flux estimations. We begin by converting diagrams of metabolic pathways into mathematical models composed of differential equations and algebraic constraints. The differential equations contain parameters related to the metabolic fluxes in the pathway of interest. Using local sensitivity analysis, we investigate the sensitivity of the model’s output to small changes in parameter values. We explore different types of dynamics caused by a large separation in timescales in the model with fast-slow analysis. Combining the knowledge discovered from these analyses we gain a deeper understanding of the limitation of parameter estimations for KFP models. The resulting estimations using the Bayesian method illustrate the challenges in estimating fast turnover rates when they are downstream of slower rates. By modeling reversible reactions, we uncover new complexities in the parameter estimation process.
- Academic Unit
- Craniofacial Anomalies Research Center; Interdisciplinary Studies Program
- Record Identifier
- 9984647452402771
Metrics
2 File views/ downloads
18 Record Views