Collaborative Research: Statistical Modeling of Mechanosensing by Cell Surface Receptors

这是佐治亚理工学院和罗格斯大学的合作团队，由两名统计学家和一名生物医学工程师组成。它可以作为一个榜样，说明如何使用严格的统计方法来解决生物学中的重要问题。这项拟议的研究将提供对称为细胞黏附的复杂生物信号过程的更好理解，这可能为未来的临床干预铺平道路。拟议的统计方法很容易适用于各种科学学科，并将对加速涉及复杂实验的众多领域的发现产生立竿见影的影响。这项研究将促进统计学和细胞生物学之间的一种新的智力互动模式。为了培养下一代数学生物学家和生物计量学家，提出了一些推广计划。该团队致力于在他们的实验室中创造一个种族、性别和民族出身的多样化环境。这项研究还将提供一个极好的机会，从代表性不足的群体中招募学生参加生物学和统计学之间的交界处的项目。细胞使用它们的表面受体通过与相邻细胞或细胞外基质上的配体结合来感知环境。这项研究的重点是了解受体-配体结合是如何诱导细胞反应的，这对于揭开许多疾病的病理机制至关重要，并可以为临床干预提供基础。除了少数例外，对于大多数受体来说，信号启动背后的机制仍然难以捉摸。该项目的创新之处在于将单分子实验与统计建模相结合，以提取理解复杂信号过程所需的新读数。基于对高斯过程(GP)模型的新改进和新的区域转换模型，提出了量化受体/配体结合中的记忆效应的新框架。为了严格量化不同触发参数对细胞信号的影响，提出了一种新的变系数Cox模型。建议的统计模型将用实验室的实验数据进行验证，并在必要时进行修改。所提出的研究具有重要意义，因为复杂的统计模型将极大地促进理解机械力对人体内两个具有生物重要性和临床相关性的受体：血小板膜糖蛋白Ib和T细胞受体的影响。从统计学的角度来看，所提出的二值数据GP模型类似于具有内插性质的标准GP模型，这可能会导致空间统计学的进一步发展。新的制度转换模型借用了不同时间序列的力量，这可能会对纵向研究产生重大影响。新的考克斯模型允许协变量的影响随时间而变化，并纳入了受试者之间的差异。它可以为研究涉及生存或失败分析的各个领域的问题开辟新的途径，并为理论和应用研究注入活力。

This is a collaborative team between Georgia Tech and Rutgers, which consists of two statisticians and one biomedical engineer. It can serve as a role model on how rigorous statistical methods are used to tackle important problems in biology. The proposed research will provide a better understanding of a complex biological signaling process called cell adhesion, which can pave the way to future clinical interventions. The proposed statistical approaches are readily applicable to a variety of scientific disciplines and will have immediate impact on accelerating discoveries in numerous fields involving complex experiments. This research will facilitate a new mode of intellectual interaction between statistics and cell biology. Some outreach programs are proposed for educating the next generation of mathematical biologists and biometricians. The team is committed to creating a diverse environment in their laboratories in terms of race, gender and national origin. The research will also provide an excellent opportunity to recruit students from underrepresented groups to participate in projects at the interface between biology and statistics. Cells use their surface receptors to sense the environment by engaging ligands on neighboring cells or in the extracellular matrix. This research focuses on understanding how receptor-ligand engagement induces cellular response, which is critical to unraveling many disease pathologies and can provide the groundwork for clinical intervention. With a few exceptions, the mechanisms behind signaling initiation remain elusive for most receptors. The innovation of this project is in combining the single-molecule experiments with statistical modeling to extract new readouts required for understanding the complex signaling processes. New frameworks based on novel modifications to Gaussian process (GP) models and new regime-switching models are proposed to quantify the memory effect in receptor/ligand binding. To rigorously quantify effects of different putative triggering parameters on cell signaling, a new varying-coefficient Cox model is proposed. The proposed statistical models will be validated with experimental data from the lab and modified if warranted. The proposed studies are significant because the sophisticated statistical modeling will greatly empower the understanding of the impact of mechanical forces on two biologically important and clinically relevant receptors in the human body: the platelet glycoprotein Ib and the T cell receptor. From the statistical point of view, the proposed GP model for binary data provides an analogy to the standard GP models with interpolation property, which can potentially lead to further advances in spatial statistics. The new regime-switching models borrow strength across different time series, which can have significant impacts on longitudinal study. The new Cox model allows the effects of covariates to vary over time and incorporates the between subject variation. It can open up new avenues for studying problems in various fields involving survival or failure analysis, and energize both theoretical and applied research.