Pattern array: in vivo mining for novel psychoactive drug discovery

描述（由申请人提供）：在某一年，美国超过四分之一的成年人患有可诊断的精神障碍，令人震惊的是，41%的12年级学生报告说他们一生都在使用非法药物。尽管精神疾病给社会和个人带来了负担，而且在精神药物研发上投入了大量资金（尽管在药物滥用方面投入的资金明显减少），但在创新精神药物方面长期存在短缺。精神科药物开发的一个主要障碍被认为是用于筛选药物治疗效果的动物模型和以靶标为中心的药物发现方法。如果潜在的病理不是局限于生物实体，而是对药物的“系统”反应，那么对特定机制干预的关注将被证明是不满意的。数据挖掘技术越来越多地用于发现癌症或毒理学反应的预测性体外系统概况。同样，以系统为基础的体内药理学取向也被认为是改变精神科药物发现的一种方式。本应用程序的目的是充分开发一种新的体内数据挖掘策略，用于精神药物的研究和开发，我们称之为模式阵列（PA）。PA的行为背景是探索性行为。在我们的实验室和其他实验室进行的广泛的行为学、药理学和行为遗传学研究表明，探索行为是1)高度遗传的，可能反映了“硬连线”的大脑系统，2)适合数学描述和高通量，3)信息丰富，每只动物产生约105个相关数据点。我们的工作假设是，药物对这种“硬连线”系统的影响也适用于算法结构和识别。我们提出的策略当然是非常规的，但它的基础是有充分的经验基础的，并在初步研究中证明是有效的。我们建议通过三个具体目标建立PA。首先，我们将开发一个来自五个主要治疗目标领域的高质量数据库：抗精神病药、抗抑郁药、抗焦虑药、滥用药物和滥用药物治疗。在每个目标区域内，表示一系列子类和机制。其次，强大的核心数据库和新药类别的经验将提供临界质量，使我们能够通过功能增强和统计实现来提高PA的功率，通用性和可靠性。第三，我们将利用PA来挖掘潜在的行为“端点”（~100,000），并识别那些最能表征药物或药物类别的端点。这些端点代表了复杂的运动模式，通过算法定义为几种动物行为学相关变量的不同组合。最后一个具体目标的结果将是为一系列具有精神活性特性的化合物提供一套体内行为“预测因子”，并为筛选新化合物提供模板。然后，PA可用于筛选与已证实的治疗方法相似的新型药物治疗方法，从而提供一种相对快速的方法来识别具有独特治疗效用的新分子实体。公共卫生相关性：用于治疗美国相当一部分可诊断的精神障碍或药物滥用患者的精神药物开发显著下降。本应用程序的目的是充分开发一种称为模式阵列（PA）的精神治疗药物的行为效应的非传统的，新颖的体内数据挖掘策略。本申请中概述的策略可用于筛选与已证实的治疗药物相似的新化合物，从而提供一种相对快速的方法来识别具有独特治疗效用的新分子实体。

DESCRIPTION (provided by applicant): Over one-quarter of the adult population in the United States suffers from a diagnosable mental disorder in a given year and an astonishing 41% of 12th graders report some lifetime use of illicit drugs. Despite the societal and personal burden that psychiatric illness presents and the substantial investment in psychiatric drug discovery (albeit significantly less in drug abuse) there is a chronic shortfall in innovative psychiatric drugs. A primary stumbling block in psychiatric drug development is thought to be in the animal models used to screen drugs for treatment efficacy and the target-centric drug discovery approach. The focus on specific mechanistic interventions will prove unsatisfactory if the underlying pathology does not rest in a restricted biological entity but rather in a `system' response to the drug. Data mining techniques are increasingly used to discover predictive in vitro system profiles for a cancer or toxicological responses. Likewise, a system-based orientation to in vivo pharmacology has been suggested as a way to transform psychiatric drug discovery. The purpose of this application is to fully develop a novel in vivo data mining strategy for psychiatric drug research and development we have termed Pattern Array (PA). The behavioral context for PA is exploratory behavior. Extensive ethological, pharmacological and behavior genetics studies in our lab and others have shown that exploratory behavior is i) highly heritable, likely reflecting `hard-wired' brain systems, ii) amenable to mathematical description and high-throughput and iii) information-rich, generating ~105 relevant data points per animal. Our working hypothesis is that the effects of drugs on this `hard-wired' system are also amenable to algorithmic structuring and identification. The strategy we are proposing is certainly unconventional, however its foundation is well-grounded empirically and shown to work in preliminary studies. We propose to establish PA via three specific aims. First, we will develop a high quality database derived from five main therapeutic target areas: antipsychotics, antidepressants, anxiolytics, drugs of abuse and drug abuse therapeutics. Within each target area a range of subclasses and mechanisms are represented. Second, the strong core database and experience with new drug classes will provide the critical mass to enable us to boost the power, generality and reliability of PA through feature enhancements and statistical implementation. Third, we will utilize PA to mine potential behavioral "endpoints" (~100,000) and identifying those that best characterize a drug or drug class. These endpoints represent complex movement patterns, algorithmically defined as different combinations of several ethologically-relevant variables. The result of this last specific aim will be to provide a set of in vivo behavioral `predictors' for a broad range of compounds with psychoactive properties and provide a template for use in screening novel compounds. PA could then be used to screen novel pharmacotherapeutics for their similarity to proven therapeutics, thus providing a relatively rapid means to identify new molecular entities with unique therapeutic utility. PUBLIC HEALTH RELEVANCE: There is a significant decline in psychiatric drug development designed to treat the considerable portion of the US population that suffers from a diagnosable mental disorder or drug abuse. The purpose of this application is to fully develop an unconventional, novel in vivo data mining strategy for the behavioral effects of psychotherapeutic drugs termed Pattern Array (PA). The strategy outlined in this application could be used to screen novel compounds for their similarity to proven therapeutics, thus providing a relatively rapid means to identify new molecular entities with unique therapeutic utility.