Postdoctoral Research Fellowships in Biology for FY 2009

这项行动为2009财政年度的NSF博士后研究奖学金提供资金。该研究金支持杰里米·布朗的一项研究和培训计划，题为“用基因组数据推断系统发生的新统计方法”。这项研究的主办机构是加州-伯克利大学，赞助科学家是约翰·休尔森贝克。DNA序列数据正在以非凡的速度收集，这为推断生物之间的进化关系（亲缘关系）提供了丰富的新信息。 理解这些关系对于回答关于生物进化的基本生物学问题以及与人类健康和福祉相关的许多应用问题（例如，传染病传播和生物多样性保护）。 然而，用如此庞大的数据集推断遗传学的问题提出了独特的分析挑战。 这项研究开发了新的统计方法，以更准确和更有效地分析基因组数据。 特别是，这项研究设计了新的DNA序列进化模型，能够灵活地将变异纳入进化力量，如突变，选择和漂移，这些变异在基因之间可能有显着差异。 这项研究还开发了评估统计方法可能误导和返回不正确的系统发育估计的指标。这项工作的培训目标包括发展统计学，计算机科学，遗传学，生物信息学和生物体生物学交叉领域的技能。 这种跨学科的观点对于利用新的基因组数据洪流解决生物学中长期存在的重要问题至关重要。 预计将改进每年用于数千项研究的分析方法，从而更好地回答一系列重要的生物学问题。

This action funds an NSF Postdoctoral Research Fellowship for FY 2009. The fellowship supports a research and training plan entitled "New Statistical Approaches to Inferring Phylogenies with Genomic Data" for Jeremy Brown. The host institution for this research is the University of California - Berkeley and the sponsoring scientist is John Huelsenbeck.DNA sequence data are being gathered at an extraordinary pace, which offers a wealth of new information for inferring the evolutionary relationships (phylogenies) between living things. Understanding these relationships is crucial both in answering basic biological questions about organismal evolution and for a number of applied problems related to human health and well-being (e.g., infectious disease transmission and biodiversity conservation). However, the problem of inferring phylogenies with such massive data sets presents unique analytical challenges. This research develops new statistical approaches to analyze genomic data with more accuracy and efficiency. In particular, this research designs new models of DNA sequence evolution that are able to flexibly incorporate variation in evolutionary forces such as mutation, selection, and drift, which can be markedly different across genes. This research also develops metrics for assessing when statistical methods are likely to be misleading and return incorrect phylogenetic estimates.The training objectives of this work include the development of skills at the intersection of statistics, computer science, genetics, bioinformatics, and organismal biology. Such an interdisciplinary view is essential to solving longstanding, important problems in biology with the new flood of genomic data. Expected are improved methods of analysis that are used in thousands of studies per year, resulting in better answers to a range of important biological questions.