Estimating Somatic Mutation Rates Using Maximum-Likelihood Methods

描述（由申请人提供）：体细胞突变是一类不可遗传的突变，对生物体的生存具有高度有害的潜力。特别是，改变DNA保真度、细胞生长速率或免疫反应逃避的体细胞突变可以产生恶性细胞（肿瘤），从而导致生育力降低，在极端情况下甚至致命。然而，尽管体细胞突变可以对多细胞生物的健康产生影响，但对体细胞突变的速率和范围知之甚少。由于体细胞突变每代发生的频率非常低，并且不可遗传，因此即使使用高通量测序技术的最新进展，它们仍然难以使用标准技术进行分析。该项目的主要目标是设计一种最大似然（ML）方法，以克服在高通量测序中测量低频突变时出现的问题。最近的研究表明，在低覆盖率基因组计划中，最大似然方法是杂合性和罕见等位基因的无偏估计。这项工作可以扩大到估计体细胞突变率从高通量测序的任何组织与已知的细胞谱系。我建议使用蒙特卡罗模拟体细胞突变来测试这个模型，体细胞突变出现在具有可变数量参数的细胞谱系中。在这些基准之后，ML框架可以应用于高突变细胞系、真核组织和癌症基因组，以确定该框架估计体细胞突变率的能力。该项目将对体细胞突变率和突变谱提供更深入的了解。此外，这个项目的结果将涉及广泛的科学和医学科学，包括细胞衰老、肿瘤发展、突变率边界的研究。此外，这里开发的模型可以确定体细胞突变分析中高通量测序的成本效益权衡，并有可能大大降低未来个体肿瘤测序可能带来的巨大成本。

DESCRIPTION (provided by applicant): Somatic mutations are a non-heritable class of mutations that have the potential to be highly detrimental to the survival of an organism. In particular, somatic mutations that modify DNA fidelity, cellular growth rates, or immune response evasion can generate malignant cells (tumors) that can lead to a reduction in fecundity or in extreme cases, lethal. Yet despite the influence that somatic mutations can have on the health of multicellular organisms, very little is known about the rate and spectrum of somatic mutations. Because somatic mutations arise at a very low frequency per generation and are non-heritable, they remain difficult to assay using standard techniques, even with the recent advances in high- throughput sequencing technology. The primary objective of this project is to design a maximum-likelihood (ML) method that can overcome the issues that arise when surveying low frequency mutations in high-throughput sequencing. Recent work has shown that maximum-likelihood methods are an unbiased estimator of heterozygosity and rare alleles in low coverage genome projects. This work can be expanded to estimate somatic mutation rates from high-throughput sequencing of any tissue with a known cell genealogy. I propose to test this model using Monte Carlo simulation of somatic mutations that arise in a cell genealogy with a variable number of parameters. After these benchmarks, the ML framework can be applied to high-mutator cell lineages, eukaryotic tissue, and cancer genomes to determine the ability of this framework to estimate somatic mutation rates. This project will provide a deeper understanding of somatic mutation rate and spectrum. In addition, the results from this project will be relevant to a broad array of scientific and medical sciences, including the study of cell aging, tumor development, the boundaries of mutation rate. Furthermore, the models developed here can determine the cost-benefit tradeoff of high- throughput sequencing in somatic mutation assays and has potential to greatly reduce the enormous costs that may come from individual tumor sequencing in the future.