Developing robust systems biology models from high-throughput data

这项工作将开发构建可靠的数学模型的方法，以便深入了解细胞内的过程。数学模型在现代生命科学研究中发挥着至关重要的作用。它们提供了一种方法来测试和发展假说，推断无法通过实验测量的细节，并找出我们对管理给定生物过程的调控机制缺乏理解的地方。系统生物学使用模型来研究生物实体之间的相互作用如何引起整个生物系统(或子系统)的功能和行为。这些系统可以是许多不同规模的，例如相互作用的生物分子组(例如蛋白质或基因)、细胞或生物体。通过用数学形式表示我们目前对生物系统的信念和假设，我们可以测试模型的行为(例如，对动态行为的模拟)是否重现了我们通过实验观察到的东西。然而，任何模型都是真实系统的抽象和简化表示；为了使建模结果有用并增加我们对生物学的理解，理解与任何结论和预测相关的不确定性是至关重要的。这项研究将开发方法来解决系统生物学研究中的两个关键需求：i)评估与建模结论相关的真实不确定性；ii)使用使用高通量技术收集的实验数据来提供细胞内过程的详细模型。这些方法将被应用于两个生物系统作为例子：i)在胚胎生长过程中确保脊髓正确发育的机制，以及ii)允许细胞感知和响应环境的蛋白质信号通路。这项工作旨在为这些特定的生物系统提供新的见解，同时开发可广泛应用于使用数学模型解决其他生物医学研究问题的方法。这项提议将扩展和修改我之前的工作，开发一种方法来测试生物系统确切结构的不确定性如何影响建模结论。在使用模型解释生物数据时，这些问题经常被忽视，但理解我们在构建模型时所做的选择可能如何影响得出的结论是至关重要的。通常有许多与我们目前的理解和数据一致的似是而非的模型，所以我们必须考虑用一个选定的模型获得的结果是否对某种程度的模型不确定性是稳健的(或敏感的)。这些发展将被应用于使用从高通量数据集获得的新信息来构建可靠的机械模型。为了研究参与协调脊髓发育的基因调控网络，我将使用单细胞转录数据，这些数据详细说明了发育过程不同阶段单个细胞中数千个基因的表达水平。我还计划使用蛋白质组数据集构建细胞类型特定信号通路的模型，预测靶向抑制的影响，蛋白质组数据包括测量各种癌细胞系中的蛋白质丰度。这项工作将在伦敦帝国理工学院理论系统生物学小组进行。方法开发将针对弗朗西斯·克里克研究所和阿斯利康的研究人员产生的新实验数据。与合作者的定期接触将确保开发的分析方法适用于计划中的应用，并构成综合实验和数学方法的一部分。

This work will develop methods to construct reliable mathematical models that provide insight into intracellular processes.Mathematical models play an essential role in modern life sciences research. They provide a way to test and develop hypotheses, infer details that cannot be measured experimentally, and identify where we lack understanding of the regulatory mechanisms governing a given biological process. Systems biology uses modelling to study how the interactions between biological entities give rise to the function and behaviour of the overall biological system (or sub-system). These systems may be of many different scales, such as groups of interacting biomolecules (e.g. proteins or genes), cells or organisms. By representing our current beliefs and assumptions about a biological system in mathematical form, we can test whether the behaviour of the model (e.g. simulations of dynamic behaviour) recreates what we observe experimentally. However, any model is an abstract and simplified representation of the real system; for modelling results to be useful and increase our biological understanding, it is crucial to understand the uncertainty associated with any conclusions and predictions.This research will develop methods to address two key needs in systems biology research: i) assessing the true uncertainty associated with modelling conclusions, and ii) using experimental data collected using high-throughput techniques to inform detailed models of intracellular processes. These methods will be applied to two biological systems as examples: i) the mechanisms ensuring correct development of the spinal cord during embryo growth, and ii) protein signalling pathways that allow cells to sense and respond to their environment. This work aims to provide novel insights into these specific biological systems, while developing methods that can be broadly applied to address other biomedical research questions using mathematical models. This proposal will extend and adapt my previous work developing a method to test how uncertainty about the exact structure of a biological system impacts modelling conclusions. Such issues are frequently overlooked when using models to interpret biological data, but it is crucial to understand how the choices we make when constructing a model may be affecting the conclusions drawn. There are generally many plausible models consistent with our current understanding and data, so we must consider whether the results obtained with a chosen model are robust (or sensitive) to some level of model uncertainty. These developments will be applied to construct reliable mechanistic models using new information gained from high-throughput datasets. To study the gene regulatory networks involved in coordinating spinal cord development, I will use single-cell transcriptomic data that detail the expression levels of thousands of genes in individual cells at different stages of the developmental pathway. I also plan to construct models of cell-type specific signalling pathways that predict the impact of targeted inhibition, using proteomic datasets comprising measurements of protein abundances in various cancer cell lines.This work will be carried out in the Theoretical Systems Biology Group at Imperial College London. Method developments will be targeted at new experimental data generated by researchers at the Francis Crick Institute and AstraZeneca. Regular engagement with collaborators will ensure that the analytical methods developed are suitable for the planned applications, and form part of an integrated experimental and mathematical approach.