Computational methods for mechanistic understanding of inter-sample variability

项目摘要 机械地理解样本间变异性的计算方法 此R21应用程序的总体目标是开发一个计算框架，该框架将允许 预测实验样本之间的生理差异。不同样品间的差异 可以根据诸如动作电位的属性在生理水平上进行分类和描述， 在基因表达等测量方面，也是在分子水平上。将可变性联系在一起 然而，以定量的方式从一个层次到另一个层次的可变性并不是一帆风顺的。在这里，通过一个 实验研究、数学建模和统计分析的创新组合，我们将 开发允许样品之间的分子水平差异转化为定量的方法 对生理差异的预测。 这项多名首席调查员的提案利用了两个主要调查员的互补专业知识。埃里克·索比博士 是心脏生理学、数学建模和计算方法方面的专家 可变性~克里斯托夫·夏尼尔博士是干细胞生物学方面的专家，将雨生能力细胞分化为特定的 单元类型和高通量方法。两个消费物价指数的共同努力将产生新的数量 数据，并将产生新的计算方法，可广泛应用于理解不同 上下文。为达致整体计划目标，我们建议： 1.收集心脏生理测量和相关离子通道、泵和 传送者。这些测量将在逐个样本的基础上进行匹配。 2.使用动态数学模型进行基于总体的模拟，以开发定量和 关于样本之间重要基因表达差异的机械性预测 转化为生理差异。 3.使用基于回归的统计方法对实验和模拟结果进行分析，并 把这两组预测联系起来。 这项探索性研究不仅可能提供对源于心肌细胞的生理学的洞察 从干细胞出发，它也可能展示一种新的计算框架，可以用于 在许多生物背景下，对样品之间的可变性进行定量处理。

Project Summary Computational methods for mechanistic understanding of inter-sample variability The overall goal of this R21 application is to develop a computational framework that will allow for the prediction of physiological differences between experimental samples. Differences between individual samples can be catalogued and described at the physiological level, in terms of properties such as action potentials, and also at the molecular level, in terms of measurements such as gene expression. Linking variability at one level to variability at another level in a quantitative manner, however, is not straightforward. Here, through an innovative combination of experimental studies, mathematical modeling, and statistical analyses, we will develop methods that allow for molecular-level differences between samples to be translated into quantitative predictions of physiological differences. This Multiple Principal Investigator proposal utilizes the complementary expertise of the two PIs. Dr. Eric Sobie is expert in cardiac physiology, mathematical modeling, and computational approaches for understanding variability~ Dr. Christoph Schaniel is expert in stem cell biology, differentiation of pluirpotent cells into specific cell types, and high-throughput methods. The combined efforts of the two PIs will generate new quantitative data and will yield new computational methods that can be applied broadly to understand variability in different contexts. To achieve the overall project goals, we propose to: 1. Collect measurements of cardiac physiology and expression of relevant ion channels, pumps, and transporters. These measurements will be matched on a sample-by-sample basis. 2. Perform population-based simulations with dynamical mathematical models to develop quantitative and mechanistic predictions regarding how differences between samples in expression of important genes are translated into physiological differences. 3. Use regression-based statistical methods to analyze the experimental and simulation results, and to relate the two sets of predictions to each other. Not only is this exploratory research likely to provide insight into the physiology of cardiac myocytes derived from stem cells, it is also likely to demonstrate a novel computational framework that can be used for quantitative treatments of variability between samples in many biological contexts.