Statistical Methods for Incomplete Data with Measurement Errors

 描述（由申请人提供）：缺失数据、删失数据和替代标记是生物医学数据分析中常见的不完整数据问题。在这个项目中，我们感兴趣的是实验，观察和遗传研究中存在缺失数据，测量误差和替代标记的统计方法。示例包括包含无应答者或缺失项目的健康调查、具有测量误差的替代标记数据等。应用可以是纵向临床试验、多水平社区研究、遗传标记、健康调查等。不完整数据可以是模型中使用的不可验证的缺失应答或作为预测因子，即缺失应答、缺失协变量和协变量测量误差。最复杂的情况是这些困难的组合，例如，缺失的响应与协变量测量误差，删失的数据与替代标志物和测量误差等。在本项目中，最终结果将是针对纵向和生存响应的两个统计包：1）MiMe：缺失数据和测量误差的统计方法，以及2）Laso：在临床事件时间替代标志物研究中纵向和生存结局的联合建模方法。将开发功能和结构方法，它们适用于许多其他领域，例如遗传标记关联研究。该项目的成果包括创新的统计方法、敏感性分析、图形方法、案例研究、软件工具和出版物。将提供R版本，高级用户可以将此版本用于与其他方法的比较研究，或自定义此版本以进行进一步扩展。第二个版本是将这项研究的工具纳入我们的在线数据分析平台Longit信息学中心。用户可以访问Longit中的许多统计软件包、模块和动态图形进行数据分析。出于各种商业化目的，我们将提供在线和离线版本，即互联网，互联网和桌面版本。我们还将授权ou API版本与商业和其他非生物医学领域的其他分析系统集成。一个例子是将Longit与Alteryx集成，Alteryx是一种用于大数据分析的商业数据挖掘工具。

 DESCRIPTION (provided by applicant): Missing data, censored data and surrogate markers are common incomplete data problems in biomedical data analysis. In this project, we are interested in statistical methods for experimental, observational, and genetic studies where there exist missing data, measurement errors, and surrogate markers. Examples include health surveys containing non-responders or missing items, surrogate marker data with measurement errors, etc. The applications could be longitudinal clinical trials, multilevel community studies, genetic markers, health surveys, etc. The incomplete data could be the non-ignorable missing response used in a model or as predictors, i.e. missing response, missing covariate, and covariate measurement errors. The most complicated scenario is the combination of such difficulties, e.g. missing response with covariate measurement errors, censored data with surrogate markers and measurement errors, etc. In this project, the ultimate results will be two statistical packages aiming at longitudinal and survival responses: 1) MiMe: statistical methods for missing data and measurement errors, and 2) Laso: joint modeling methods for longitudinal and survival outcomes in the study of surrogate marker for clinical event time. Functional and structural approaches will be developed, and they are applicable to many other areas, e.g. genetic markers association studies. The results from this project include innovative statistical methods, sensitivity analysis, graphical methods, case studies, software tools, and publications. An R version will be available and advanced used may apply this version for comparison studies vs. other approaches or customize this version for further extensions. A second version is to incorporate the tools from this research into our online data analysis platform, the Longit Informatics Center. Subscribers can access many statistical packages, modules, and dynamic graphics in Longit for data analysis. For various commercialization purposes, we will deliver online and offline versions, i.e. internet, intraweb, and desktop versions. We will also license ou API version for integrating with other analytic systems in business and other non-biomedical fields. One example is to integrate Longit with Alteryx, a commercial data mining tool for big data analysis.