In Silico Screening of Medications for Slowing Alzheimer's Disease Progression.

减缓阿尔茨海默病进展药物的计算机筛选。

• 批准号: 9884696
• 负责人: RICHARD E KENNEDY
• 金额: $ 64.6万
• 依托单位: UNIVERSITY OF ALABAMA AT BIRMINGHAM
• 依托单位国家: 美国
• 财政年份: 2017
• 资助国家: 美国
• 起止时间: 2017-09-15 至 2022-03-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/9884696
• 关键词: Algorithms; Alzheimer' s Disease; Alzheimer' s disease diagnosis; Alzheimer' s disease therapy; Bioinformatics; Clinical Data; Clinical Trials; Cognitive; Combination Medication; Complement; Data; Data Set; Databases; Dementia; Development; Diagnosis; Disease; Disease Progression; Drug Screening; Electronic Health Record; Failure; Future; Gene Expression; Gene Targeting; Genes; Goals; Individual; Investments; Literature; Malignant Neoplasms; Measures; Medical; Medicare; Medicare claim; Methods; Modeling; Modification; Natural experiment; Outcome; Participant; Patients; Pattern; Pattern Recognition; Pharmaceutical Preparations; Phase; Process; Randomized Controlled Trials; Research; Resources; Rewards; Sampling; Series; Source; Testing; Therapeutic; Therapeutic Agents; Therapeutic Effect; Time; United States National Institutes of Health; Validation; base; beneficiary; clinical Diagnosis; clinical care; clinical encounter; cognitive performance; cognitive testing; cohort; cooperative study; cost; data mining; drug development; epidemiology study; follow-up; improved; in silico; in vivo; interest; large scale data; mild cognitive impairment; multidimensional data; neuroimaging; novel; novel strategies; novel therapeutics; screening; symposium; text searching; therapy development

Drug development in Alzheimer's disease (AD) requires a considerable investment of time and re- sources, often with little reward as the vast majority of medications ultimately prove unsuccessful. Drug repur- posing, in which medications that already have been approved for treatment are evaluated for therapeutic effects in other disorders, has the potential to markedly increase the number of agents in the drug development pipeline but requires methods for effective screening of candidate medications for activity. In silico or computational ap- proaches to medication screening are rapidly growing, and have been successful in illnesses such as cancer, but their application to AD remains understudied. There is also intense interest in drug repurposing approaches that will utilize the vast amounts of clinical data that are being collected from epidemiological studies and clinical encounters documented through electronic health records (EHRs). In this proposal, we present a novel approach to drug repurposing that uses large-scale data mining (i.e., pattern recognition) algorithms applied to concurrent medication taken by participants in AD clinical trials and in Medicare administrative data to determine which of these medications show potential therapeutic benefits. With over 30 years of AD clinical trial data available to us through a recently developed meta-database and 10 years of prescription data available through Medicare Part D, the administration of concurrent medications to patients as part of their routine clinical care constitutes a large- scale natural experiment. This information can be harnessed for AD treatment discovery if appropriate methods can be developed to detect effects on disease progression within this high-dimensional data. Data mining al- gorithms that discover patterns of associations in data, rather than testing predetermined hypotheses, are well suited to application in large-scale screening for drug repurposing. Using our meta-database and Medicare data, we will be able to evaluate most of the more than 6,000 currently available prescription medications for efficacy in AD using well-accepted endpoints for measuring disease progression. The discovery phase will be followed by a validation phase of promising candidate medications in independent data sets, as well as identification of plausible gene targets for each medication from the biomedical literature. This study will set the groundwork for a series of follow-up in vivo studies to conclusively demonstrate effects of selected medications for AD, expanding the current armamentarium for treating this common and debilitating disorder.

治疗阿尔茨海默病(AD)的药物开发需要投入大量的时间和精力 消息来源，往往没有什么回报，因为绝大多数药物最终被证明是不成功的。毒品声誉- 姿势，其中已经被批准治疗的药物被评估治疗效果 在其他疾病中，有可能显著增加药物开发流水线中的药物数量 但需要有效筛选活性候选药物的方法。在计算机或计算应用程序中- 药物筛查的途径正在迅速增长，并在癌症等疾病中取得了成功， 但它们在AD中的应用仍未得到充分研究。人们对药物再利用的方法也非常感兴趣。 这将利用从流行病学研究和临床中收集的大量临床数据 通过电子健康记录(EHR)记录的接触。在这个提案中，我们提出了一种新的方法 使用大规模数据挖掘(即模式识别)算法应用于并发的药物再利用 AD临床试验和联邦医疗保险管理数据中参与者服用的药物，以确定 这些药物显示出潜在的治疗效果。有超过30年的AD临床试验数据提供给我们 通过最近开发的元数据库和Medicare Part提供的10年处方数据 作为患者常规临床护理的一部分，同时给患者用药构成了一个很大的- 规模化自然实验。如果适当的方法，可以利用这些信息来发现AD的治疗方法 可以开发成在这些高维数据中检测对疾病进展的影响。数据挖掘等- 发现数据中关联模式的算法，而不是测试预先确定的假设的算法是很好的 适用于药物再利用的大规模筛选。使用我们的元数据库和医疗保险数据， 我们将能够评估目前可用的6,000多种治疗EFfiCacy的处方药中的大多数 在AD中，使用公认的终点来衡量疾病进展。接下来是发现阶段 通过独立数据集中有希望的候选药物的验证阶段，以及身份fi阳离子 生物医学文献中每种药物看似合理的基因靶点。这项研究将为 一系列活体跟踪研究，最终证明选定的药物治疗AD的效果，范围扩大 目前治疗这种常见的、使人衰弱的疾病的设备。