Genome-wide mutation models to decipher function

描述（由申请人提供）：完整的脊椎动物基因组正在迅速积累，积累的步伐只会加快。这是一个好消息，因为比较分析的效用在很大程度上取决于物种采样的多样性。然而，要充分利用如此广泛的数据的潜力，还面临着巨大的挑战：需要开发新的方法和分析方法。我们的目标是发展和扩大我们的能力，分析动态进化过程（跨地区和通过时间），形成了现存的基因组。我们将使用我们最近开发的贝叶斯进化分析方法来实现这一目标，该方法使我们比竞争方法具有许多数量级的速度优势，并且可以很好地扩展模型复杂性和数据大小。我们提出的许多研究都是基于生物现实的范例，以前是不可能考虑或测试，因为计算的限制。我们建议全面描绘一组选定的脊椎动物基因组的重复内容，包括从基因组的“暗物质”（目前未注释的部分）的古老元素的注释。这组重复序列中的转座因子将用于构建第一个完整的全基因组上下文相关替代过程模型。我们将考虑重组，重排，表达和局部核苷酸含量等背景，以及未知的背景，并分析受这些背景影响的进化过程如何随着时间的推移而变化。这些上下文依赖的替代模型将为脊椎动物基因组种间比较中识别和注释功能区域以及区分和表征蛋白质中基于适合度的效应提供强有力的工具。核心概念是，如果我们更好地了解背景核苷酸取代的全基因组模式，那么我们将能够更准确地识别可能具有功能的基因组区域，并了解选择如何指导蛋白质的进化。

DESCRIPTION (provided by applicant): Complete vertebrate genomes are accumulating rapidly, and the pace of accumulation will only increase. This is excellent news, because the utility of comparative analysis depends heavily on the diversity of species sampling. There are, however, substantial challenges to exploiting the full potential of such extensive data: development of novel methods and analytical approaches is needed. We aim to develop and extend our capacity to analyze the dynamic evolutionary processes (across regions and through time) that have shaped extant genomes. We will achieve this goal using a Bayesian evolutionary analysis approach we recently developed that allows us many orders of magnitude speed advantage over competing approaches, and which scales well with model complexity and data size. Many of the studies we propose are based on biologically realistic paradigms that previously were impossible to consider or test because of computational limitations. We propose to comprehensively delineate the repetitive contents of a selected set of vertebrate genomes, including annotation of ancient elements from the "dark matter" of genomes (the currently unannotated portion). The transposable elements in this set of repeat sequences will be used to build the first complete genome-wide models of context-dependent substitution processes. We will consider contexts such as recombination, rearrangement, expression, and local nucleotide content, as well as unknown contexts, and analyze how the evolutionary processes influenced by these contexts have changed over time. These context- dependent substitution models will provide a powerful tool for identifying and annotating functional regions in interspecific comparisons of vertebrate genomes, and for differentiating and characterizing fitness-based effects in proteins. The core concept is that that if we better understand genome-wide patterns of background nucleotide substitution, then we will be able to more accurately identify genomic regions that are likely functional, and to understand how selection directs the evolution of proteins.