Large-Scale Methods for Assessing the Consequences of Mutations in Proteins

描述（由申请人提供）：当代基因组学的基本挑战之一在于理解基因组改变如何产生疾病。由于几个因素，迎接这一挑战的紧迫性日益增加。首先，我们已经了解到，每个个体都拥有数量惊人的罕见蛋白质编码变体，其功能后果将难以使用基于关联的方法来解决。其次，我们在了解与许多疾病有关的基因和途径方面取得了令人难以置信的进步。因此，我们非常接近于能够为医生、病人和基因测试的临时用户提供个性化的、基于基因组的建议。然而，由于缺乏有效的方法来确定我们在典型人类基因组的蛋白质编码区发现的约300种罕见变异的功能后果，我们受到了阻碍。目前评估罕见蛋白质编码变异后果的方法要么是实验性的，要么是计算性的。实验方法通常包括对蛋白质功能的细胞或生化分析。虽然这些方法是有效的，但它们是在个案的基础上使用的，不能扩展到解决我们在每个人类基因组中发现的罕见变异。用于确定蛋白质变异影响的计算方法虽然易于扩展，但通常会产生大量的假阳性和阴性结果。因此，需要一种新的方法来研究蛋白质编码变异的功能后果。我们建议通过开发方法来直接测量所有可能的单个突变的功能后果，同时使用真核模型系统来解决这一需求。我们可以利用这些数据来创建疾病相关蛋白的序列功能图谱，这将使更有效的基因诊断成为可能。为了实现这一目标，我们将利用我们的专业知识，将蛋白质功能测定与高通量DNA测序相结合，同时测量数十万种蛋白质变体的功能后果。此外，我们将通过同时研究诱变对多种细胞表型的影响，开始剖析突变对蛋白质的复杂性。

DESCRIPTION (provided by applicant): One of the fundamental challenges in contemporary genomics lies in understanding how genomic alterations produce disease. An increasing urgency to meet this challenge has arisen owing to several factors. First, we have learned that every individual harbors a surprisingly large number of rare, protein-coding variants whose functional consequences will be difficult to address using association-based methods. Second, we have made incredible strides in understanding the genes and pathways involved in many diseases. As a result, we are tantalizingly close to being able to offer personalized, genomically-based advice to physicians, patients and casual users of genetic tests. However, we are hampered by our lack of effective methods for determining the functional consequences of the ~300 rare variants we find in the protein-coding regions of a typical human genome. Current methods for assessing the consequences of rare protein-coding variants are either experimental or computational. Experimental methods generally involve cellular or biochemical assays for protein function. Though these methods are effective, they are used on a case-by-case basis, which cannot be scaled to address the rare variants we find in each human genome. Computational methods for determining the impact of protein variants, though easily scalable, generally produce a large number of false positive and negative results. Thus, a novel approach to studying the functional consequences of protein-coding variation is needed. We propose to address this need by developing methods for directly measuring the functional consequences of all possible single mutations in a protein simultaneously using eukaryotic model systems. We can use these data to create sequence-function maps for disease-related proteins, which will enable more effective genetic diagnosis. To accomplish this goal, we will draw on our expertise in combining assays for protein function with high-throughput DNA sequencing to measure the functional consequences of hundreds of thousands of variants of a protein simultaneously. Furthermore, we will begin to dissect the complexity of mutational effects on proteins by studying the impact of mutagenesis on multiple cellular phenotypes simultaneously.