Method for Prediction of Efficacy of Genetic-Based Prediction Models of Personali

描述(由申请人提供)：新发现的具有临床价值的基因测试的集合远远超过了我们使用临床试验来确定其临床有效性并确定哪些测试集合和相关算法最适合任何给定临床设置的能力。例如，对文献的回顾表明，在过去的十年中，至少有35种已发表的华法林治疗性剂量预测算法被发表，其中9种算法包含了基因数据(未发表的综述)。确定这些算法中最有临床用途的支出包括总共19项临床试验(Clinicaltrials.gov，2008年2月访问)。使这种复杂性变得更加复杂的是，缺乏将这些发现成功转化为临床应用所需的临床和生物医学信息学方法、工具和基础设施，这加剧了迅速扩大的临床有价值的基因发现集合的成功医疗应用的延迟。将重要的生物医学信息学方法、工具和过程转化为在临床环境中实施重要的新基因发现所需的努力，受到监管、技术和验证障碍的严重阻碍，这些障碍在当前的临床研究或临床企业环境中不易解决。这项提案将测试创建临床化身的环境和方法，这些化身的数据在统计上与实际患者的电子病历一致。一旦创建，这些临床化身医疗记录将被用于进行Insilo实验，以比较基于遗传的算法及其在临床环境中的预测值。我们将开发和测试该方法，并通过进行一系列临床化身模拟和随后基于遗传的华法林剂量预测分析来创建应用程序。一旦开发出来，我们将进行一系列Insilo临床试验，比较各种临床和遗传因素，以证明华法林剂量算法的有效性。这项工作是类似项目的代表，这些项目可能旨在测试其他个性化医疗设备，例如用于量化癌症风险的基因测试、药物遗传学问题，以及其他FDA标记的“IVDMIA”设备(例如，Genetic Health的Oncotype DX(R))，预计将改善预测性和预防性个性化医疗。

DESCRIPTION (provided by applicant): The collection of newly discovered clinically valuable genetic tests is far outpacing our ability to use clinical trials to determine their clinical efficacy and determine which of the collection of tests and associated algorithms are best suited for any given clinical setting. For example, a review of the literature demonstrates that at least 35 published algorithms for the prediction of therapeutic warfarin dosing have been published in the past ten years, nine of which include genotype data (unpublished review). Expenditures to determine the most clinical useful of these algorithms include a total of 19 clinical trials (clinicaltrials.gov, accessed Feb, 2008) Compounding this complexity and adding to the delay of the successful medical use of the rapidly expanding collection of clinically valuable genetic discoveries is the lack of clinical and biomedical informatic methods, tools and infrastructure required to facilitate the successful translation of the discoveries to practical clinical use. Efforts to translate important biomedical informatics methods, tools and processes required to implement important new genetic discoveries in the clinical setting are severely hindered by regulatory, technical and validation barriers not easily resolved in the current clinical-research or clinical enterprise environments. This proposal will test an environment and methodology that creates clinical avatars with data statistically consistent with actual patient electronic medical records. Once created, these clinical avatar medical records will be used to conduct insilico experiments to compare genetic- based algorithms and their predicted value in the clinical setting. We will develop and test the methodology and create the applications by conducting a series of clinical avatar simulations and subsequent analysis of genetic-based warfarin dosing prediction. Once developed, we will conduct a series of insilico clinical trials comparing various clinical and genetic factors to demonstrate the efficacy of the warfarin dosing algorithms. This work is representative of similar projects that may be designed to test other personalized medicine devices such as genetic tests used to quantify risk to cancers, pharmacogenetic problems, and other FDA labeled "IVDMIA" devices (e.g. Genetic Health's Oncotype DX(R)) projected to improve predictive and preventive personalized medicine.