Statistical and Computational Tools for Identifying Microarray-based Expression and Single Feature Polymorphism Markers in Crops

加州大学河滨分校（University of California, Riverside）被授予开发统一的自动化统计算法的奖项，该算法用于在作物基因组尺度上发现新开发的基于微阵列的单特征多态性标记（SFPs）和基因表达标记（GEMs）。该研究所成功开发了基于鲁棒投影追踪（robust projection pursuit， RPP）方法的两种基因型间SFPs识别算法，并应用于大麦、水稻、豇豆和小麦基因组。最近，一种新型的基于微阵列的基因表达标记（GEMs）被开发出来，这需要一种统计上可靠的识别方法。该项目旨在扩展RPP方法，以便在多基因型中更有效地鉴定SFPs（目标1）；构建基于混合模型的GEMs识别方法（Aim 2）；使用表达的序列标签和直接测序来验证检测到的SFPs和GEMs（目的3）；结合SFPs和GEMs进行基因分型，在期望最大化算法框架内嵌入隐马尔可夫模型（HMM）构建遗传图谱（Aim 4）；制定一项外联计划，包括为作物研究和育种界提供培训和研究经验（目标5）。这些工具将通过使用大量公开可用的拟南芥和大麦表达数据来证明。该项目将为作物界提供新的资源，以生产丰富的、高通量的功能相关标记，用于在基因组尺度上详细分析特定性状的遗传基础，了解属间的基因组差异，并为实现迫切需要的品种改良目标提供新的途径。该项目将通过吸引女性和少数族裔申请的校园项目扩大参与范围。该项目的跨学科性质将促进我们实验室内外生物学家，计算机科学家和统计学家之间的合作，这些互动将加强UCR遗传学，基因组学和生物信息学（GGB）的新兴研究生课程。

The University of California, Riverside, is given an award to develop unified, automated statistical algorithms to discover newly-developed microarray-based single-feature polymorphism markers (SFPs) and gene expression markers (GEMs) in crops on a genome scale. The PI has successfully developed the algorithm based on robust projection pursuit (RPP) method to identify SFPs between two genotypes, with application in Barley, Rice, Cowpea and Wheat genome. Recently, a new type of microarray-based gene expression markers (GEMs) has been developed that calls for a statistically sound identification method. The project aims to extend the RPP method for more powerful SFPs identification among multiple genotypes (Aim 1); to build a mixture model-based method for GEMs identification (Aim 2); to validate the detected SFPs and GEMs using expressed sequence tags and direct sequencing (Aim 3); to combine SFPs and GEMs for genotyping and construct genetic map by embedding hidden Markov model (HMM) within the framework of the expectation-maximization algorithm (Aim 4); to develop an outreach program that includes training and research experiences for crop research and breeding community (Aim 5). The tools will be demonstrated by using a large publicly available Arabidopsis and Barley expression data. This project will provide the crop community with new resources for producing abundant, high throughput functional-related markers for detailed analysis of the genetic basis of specific traits on a genomic scale, for understanding genomic divergence between genera, and for implementing new approaches toward critically needed cultivar improvement objectives. The project will broaden participation using campus programs that attract female and minority applications at UCR. The inter-disciplinary nature of the project will promote collaborations within and outside our laboratory between biologists, computer scientists and statisticians, and these interactions will strengthen an emerging graduate program in Genetics, Genomics and Bioinformatics (GGB) at UCR.