Mathematical and statistical modelling of cell fate in Haematopoiesis

造血细胞命运的数学和统计模型

• 批准号: 2271097
• 金额: --
• 依托单位: University of Oxford
• 依托单位国家: 英国
• 项目类别: Studentship
• 财政年份: 2019
• 资助国家: 英国
• 起止时间: 2019 至 无数据
• 项目状态: 已结题
• 来源: https://gtr.ukri.org/projects?ref=studentship-2271097
• 关键词: Mathematical; statistical; modelling; cell; fate

The way in which blood stem cells first emerge and go on to function in full homoeostasis is not yet fully understood. This project will address this question via mathematical modelling applied to lineage tracing and single cell data. The student will develop stochastic and Bayesian models to model cell fate dynamics in haematopoiesis from the point that the blood stem cells are first produced to the steady state dynamics that follow it. The data used to calibrate the model will be lineage tracing data, of which there are a number of public data sets for adult mouse and the De Bruijn lab is currently working on for development. Once the model is built it will be used in two ways, first to assess the effect of mutations with lineage tracing data where a mutation is induced along with the lineage tracing label. Secondly Bayesian models will be developed to combine the model with single cell data to study the fate mechanisms at a molecular level. During the course of the PhD, the student will develop skills in probabilistic modelling, stochastic process and deep learning, as well as knowledge of the haematopoietic system. The project is both quantitative and interdisciplinary. The candidate comes from a pure maths background and will be trained in the use of stochastic modelling and statistical machine learning (including Bayesian Inference and deep learning) applied to studying cell fate and homeostasis in haematopoiesis. The project is fully interdisciplinary, with one supervisor a former theoretical physicist that works on modelling stem cells and the other a biomedical scientist with extensive expertise in blood stem cells.

造血干细胞最初出现并在完全体内平衡中发挥作用的方式尚未完全了解。该项目将通过应用于谱系追踪和单细胞数据的数学建模来解决这个问题。学生将开发随机和贝叶斯模型来模拟造血中的细胞命运动态，从造血干细胞第一次产生到随后的稳态动态。用于校准模型的数据将是谱系追踪数据，其中有许多成年小鼠的公共数据集，De Bruijn实验室目前正在开发。一旦建立了模型，它将以两种方式使用，首先是评估具有谱系追踪数据的突变的影响，其中突变是沿着谱系追踪标签诱导的。其次，将开发贝叶斯模型以将该模型与单细胞数据联合收割机结合以在分子水平上研究命运机制。在博士课程期间，学生将发展概率建模，随机过程和深度学习的技能，以及造血系统的知识。该项目是定量和跨学科的。候选人来自纯数学背景，将接受随机建模和统计机器学习（包括贝叶斯推理和深度学习）的培训，用于研究造血中的细胞命运和稳态。该项目是完全跨学科的，一位主管是前理论物理学家，致力于干细胞建模，另一位是在血液干细胞方面具有广泛专业知识的生物医学科学家。