CORE--STATISTICAL ANALYSIS OF TOXICS MEASUREMENT

统计核心的主要作用将是就实验设计提供咨询，并协助研究项目和其他核心进行数据解释。我们还将向项目提供数据库支持，特别注意对分析核心A和DNA微阵列核心C的支持。新的重点将放在生物信息支持领域。我们将开发存储和解释DNA微阵列数据的方法，并协助模式识别系统与模式识别系统进行蛋白质组和总基因组学方法的模式识别。此外，我们还将为新兴的微阵列领域开发新的统计方法。我们将继续为免疫阵列和其他高密度信息系统开发统计方法。将采用贝叶斯估计、非线性最优设计、最大似然估计和M估计等技术。我们建议开发和改进用于危险化学品分析的统计方法和算法，并创建新的和改进的测量技术。对于加速器质谱学的数据分析，这一点尤为重要。我们将处理包括非线性校准、非恒定方差、可能的异常值、接近或低于检测下限的值以及高维和大数据集等问题。这些目标将通过五个具体目标实现，包括：在统计问题上与其他调查人员合作，为危险化学品研究中常见的问题开发新的统计方法，在数值分析和数值优化中应用最先进的技术，开发高效、快速和可靠的计算机算法，以解决实施统计方法方面的问题，开发用于实验室的软件，纳入统计方法和数值方法的上述发展，协助解决复杂结构的数据文件引起的生物信息学问题，以及开展外联活动，将新的统计技术转让给环境和监管界。

The primary role of the statistics core will be to provide consultation on experimental design and assistance with data interpretation for the research projects and other cores. We also will provide database support to projects with particular attention to support of the Analytical Core A and the DNA microarray Core C. A new emphasis will be in the area of bioinformatic support. We will develop methods for the storage and interpretation of DNA microarray data and assist with pattern recognition systems with pattern recognition systems for proteomic and metanomic approaches. In addition we will develop new statistical methods for the emerging field of microarrays. We will continue our work on developing statistical methods for immunoarrays and other information systems with high density. Techniques such as Bayes estimation, non-linear optimal design, maximus pseudo likelihood, and M estimation will be employed. We propose the development and refinement of statistical methods and algorithms for the analysis of hazardous chemicals and for the creation of new and improved measurement techniques. This will be particularly important with the analysis of data from accelerator mass spectroscopy. We will deal with problems including non-linear calibration, non-constant variance, possible outliers, values near or below detection limits and high- dimensional and large data sets. These goals will be accomplished in five specific aims including: collaboration with other investigators with other investigators on statistical questions, develop new statistical methods for problems commonly encountered in hazardous chemical research, apply state-of-the-art techniques in numerical analysis and numerical optimization to develop efficient, fast, and reliable computer algorithms to solve problems in implementing statistical methods, develop software incorporating the above developments in statistical methodology and numerical methods for laboratory use, assist with bioinformatic problems induced by the complexly structured data files, and conduct outreach activities to transfer new statistical technology to the environmental and regulatory community.