Mechanistic models of cardiac action potentials and drug effects in rabbit ventricular myocytes

Yang, Zhechao (2026) Mechanistic models of cardiac action potentials and drug effects in rabbit ventricular myocytes. PhD thesis, University of Glasgow.

Abstract

The heart’s rhythmic electrical activity is governed by a complex interplay of ionic currents, which vary significantly between cells and individuals. Accurately modelling this variability is crucial for understanding normal cardiac function and predicting the impact of therapeutic interventions. This thesis presents three complementary studies in computational cardiac electrophysiology, focusing on parameter sensitivity, intercellular variability, and pharmacodynamic modelling in rabbit ventricular myocytes.

First, a global Sobol sensitivity analysis was applied to the Shannon model of rabbit ventricular action potentials to determine the ionic currents most influential in shaping key electrophysiological biomarkers. This analysis identified the background chloride current (IClb) as the primary driver of variability, with substantial contributions from IK1, IKr, IKs, INaCa, Itos, and ICaL. A reduced model retaining just six conductances was shown to reliably reproduce action potential duration and other biomarkers across varying conditions, streamlining future personalization efforts.

In the second study, fluorescence voltage recordings from over 1,200 rabbit ventricular myocytes provided by Professor Godfrey Smith from University of Glasgow were used to generate a population of cell-specific action potential models. Each model was individually fitted to replicate the full waveform of a single cell. This approach captured the natural variability in electrophysiological properties across the population and revealed that action potential metrics such as duration are maintained through complex interactions among weakly correlated ionic parameters. The results demonstrate the feasibility of high-throughput, cell-level model calibration and underscore the importance of considering inter-cellular variability in cardiac simulations.

Finally, the third project explored the pharmacodynamics of the antiarrhythmic drug dofetilide. By integrating paired recordings of action potentials before and after drug application, a novel model-based framework was developed to test the predictive accuracy of inferred conductances. Dofetilide-induced changes in action potential duration were simulated by scaling the rapid delayed rectifier potassium current (IKr), and predictions were validated across four drug concentrations. The results demonstrated strong model performance at lower concentrations and highlighted increasing variability at higher doses, underscoring the need for more comprehensive models at therapeutic extremes.

Together, these studies emphasize the importance of accurate parameter identification, the role of cellular heterogeneity in shaping electrophysiological behavior, and the value of mechanistic modelling in drug safety assessment. This work advances the field toward personalized cardiac modelling and improves our ability to predict both natural and drug-modified electrical dynamics in the heart.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Additional Information: Supported by funding from the SofTMech team.
• Subjects: Q Science > QA Mathematics
• Colleges/Schools: College of Science and Engineering > School of Mathematics and Statistics > Mathematics
• Funder's Name: SofTMech team
• Supervisor's Name: Simitev, Professor Radostin and Gao, Dr. Hao
• Date of Award: 2026
• Depositing User: Theses Team
• Unique ID: glathesis:2026-86096
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 14 Jul 2026 11:12
• Last Modified: 14 Jul 2026 11:13
• Thesis DOI: 10.5525/gla.thesis.86096
• URI: https://theses.gla.ac.uk/id/eprint/86096

Related URLs

• Article DOI
• Article DOI