STRUCTURAL VISUALIZATION IN BIOINFORMATICS

这个子项目是许多研究子项目中利用 资源由NIH/NCRR资助的中心拨款提供。子项目和 调查员(PI)可能从NIH的另一个来源获得了主要资金， 并因此可以在其他清晰的条目中表示。列出的机构是 该中心不一定是调查人员的机构。 自上次报告所述期间以来，关键人员的支助没有变化。估计的未承付余额不会超过当年预算的25%。 A.具体目标： 这笔赠款的具体目的与原提案中所述相同。 B.今年的进展情况： 赠款最后一年的大部分工作集中于具体目标1D、2A和2B。具体目标1D是指为提取模式而提供的对微阵列数据的平台范围的分析。具体目标2A和具体目标2B是指开发基于网络的软件工具，以进行结构分析和可视化。 我们基于微阵列样本的累积分布函数(CDF)开发了几种分析技术，以检测在所有实验中典型的离群值、不寻常的表达模式和表达模式。这些方法是结构化分析技术，因为它们以典型的平台范围的CDF为背景来分析CDF。我们开发了一种计算“基因效应”的方法，它是独立于实验条件的与每个基因相关的基线表达水平。这些影响只能通过查看大量样本来计算。具有高基因效应的基因很可能是家政基因，因此基因效应模型提供了一种确定家政基因的独立方法。在寻找高表达的基因时，表达的新颖性在某种程度上取决于其基因效应的价值。我们已经证明，基因效应在不同平台上是一致的。这些结果已经在几份技术报告和正在修订的两篇论文中得到了报道。我们正在努力证明基因效应在差异表达分析中发挥着重要作用，并将这些结果应用于几个大规模的癌症微阵列研究。

This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. The subproject and investigator (PI) may have received primary funding from another NIH source, and thus could be represented in other CRISP entries. The institution listed is for the Center, which is not necessarily the institution for the investigator. There has been no change in the support of key personnel since the last reporting period. The estimated unobligated balance will not be greater than 25% of the current year's budget. A. Specific Aims: The specific aims of this grant are as stated in the original proposal. B. Progress this year: The majority of the work in the final year of the grant was focused on Specific Aims 1D, 2A and 2B. Specific aim 1D refers to provided platform-wide analysis of microarray data in order to extract patterns. Specific aim 2A and Specific aim 2B refer to the development of web-based software tools to do structural analysis and visualization. We have developed several analysis techniques based on the cumulative distribution functions (CDF) of the microarray samples to detect outliers, unusual expression patterns, and expression patterns that are typical across all experiments. These methods are structural analysis techniques in that they analyze the CDFs against the background of the typical platform-wide CDF. We developed a method of calculating a "gene effect" which is a baseline expression level associated with each gene independent of experimental condition. These effects can only be calculated by looking at a large number of samples. Genes with high gene effect are likely to be housekeeping genes and so gene effect models provide an independent method of determining housekeeping genes. When looking for genes that are highly expressed, the novelty of the expression is determined to some extent by the value of its gene effect. We have shown that gene effects are consistent across platforms. These results have been reported in several technical reports and in 2 papers that are undergoing revision. We are working to show that gene effects play an important role in the analysis of differential expression and are applying these results in several large scale cancer microarray studies.