Stochastic modelling and inference of cell cycle control on lineage trees
File(s)
Author(s)
Hughes, Fern
Type
Thesis
Abstract
Single-cell imaging technologies and lineage tracing allow for many generations of cellular observations to be captured on lineage trees. A key question is how to use the cell-to-cell variability that propagates across the tree to uncover the mechanisms underlying the cell cycle. Using the lineage tree structure, correlations in cell cycle measurements between different pairs of related cells can be calculated. It is presently unclear what information correlation patterns carry about cell cycle dynamics.
In this thesis, we propose a stochastic model which describes how the inheritance of cell cycle factors on lineage trees produces complex correlation patterns. We also incorporate this model within an agent-based framework to derive expressions that relate interdivision time heterogeneity and correlation patterns to population fitness. We use Bayesian inference to interpret the correlation patterns of both interdivision time and cell cycle phases across datasets of bacteria, fibroblasts and cancer.
Our analysis reveals that correlation patterns across a range of cell types arise from biological oscillators coupled to cell division. This rhythmicity is observed across both interdivision time correlations and correlations between individual phases of the cell cycle. We also demonstrate that fluctuations in cell interdivision times have a direct effect on the population growth rate, but the interpretation changes based on how the interdivision times are experimentally measured, which may lead to counter-intuitive effects.
This thesis provides a general framework to interpret mechanisms of cell cycle control with non-invasive methods and is applicable to single-cell observations from live imaging experiments. The predictions from the inference and theory could be used to draw testable hypotheses on the drivers of the cell cycle and cell proliferation, which could inform future experiments, discriminate between possible mechanistic models, and compare cell lines under different conditions where the effect of the condition on the cell cycle is not known.
In this thesis, we propose a stochastic model which describes how the inheritance of cell cycle factors on lineage trees produces complex correlation patterns. We also incorporate this model within an agent-based framework to derive expressions that relate interdivision time heterogeneity and correlation patterns to population fitness. We use Bayesian inference to interpret the correlation patterns of both interdivision time and cell cycle phases across datasets of bacteria, fibroblasts and cancer.
Our analysis reveals that correlation patterns across a range of cell types arise from biological oscillators coupled to cell division. This rhythmicity is observed across both interdivision time correlations and correlations between individual phases of the cell cycle. We also demonstrate that fluctuations in cell interdivision times have a direct effect on the population growth rate, but the interpretation changes based on how the interdivision times are experimentally measured, which may lead to counter-intuitive effects.
This thesis provides a general framework to interpret mechanisms of cell cycle control with non-invasive methods and is applicable to single-cell observations from live imaging experiments. The predictions from the inference and theory could be used to draw testable hypotheses on the drivers of the cell cycle and cell proliferation, which could inform future experiments, discriminate between possible mechanistic models, and compare cell lines under different conditions where the effect of the condition on the cell cycle is not known.
Version
Open Access
Date Issued
2023-06-24
Date Awarded
2024-02-01
Copyright Statement
Creative Commons Attribution NonCommercial Licence
License URL
Advisor
Thomas, Philipp
Barr, Alexis
Publisher Department
Mathematics
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)