Novel tool for linking drugs and phenotypes

描述（由申请人提供）：将新药推向市场是一个漫长而昂贵的过程。为了使消费者能够以合理的成本获得更多可负担的新药，必须大大提高药物开发过程的效率。我们的目标是开发一种新的工具，通过利用我们新开发的计算方法快速识别任何给定临床表型的药物靶点和候选药物来解决这一挑战。该方法基于一个假设，即药物靶点表达的组织模式对于表型与潜在药物的无偏匹配至关重要。因此，将疾病的复杂分子生物标志物与药物相匹配取决于对药物或药物组合的完整分子特征（多药理学指纹）的了解。我们的方法是基于整合来自化学和“组学”数据库的大型数据集，并开发将临床表型与药物和类药物化学物质联系起来的算法。这是通过识别生物活性药物的多药理学特征，并将这种基于药物的特征与基于表型的疾病组织特征相匹配来实现的，这是来自“组学”数据的计算分析。我们已经开发并验证了该系统的“药物”模块。在这里，我们建议(a)开发“表型”模块，该模块将通过整合临床表型、组织特异性靶标表达、GWAS和微阵列数据，包括表型-组织-靶标数据；(b)为模块开发“匹配”算法，在药物和临床表型之间建立直接联系，并为这些匹配建立排名系统；(c)使用成熟的临床表型和治疗药物验证该系统。一旦完成，拟议的研究将产生一个系统，该系统将允许用户立即识别任何给定临床表型的预测药物排名列表。或者，它可用于识别与任何给定药物/化合物或药物/化合物组合相关的表型列表。这解决了药物开发过程中的一个主要挑战，提高了药物发现阶段的生产力和效率，它为制药工业和学术研究提供了一个有用的和高度期望的工具。最终，这项技术将有助于理解基因组对疾病生物学的贡献，并加速基因组学的使用，以推进医学科学和医疗保健的有效性。

DESCRIPTION (provided by applicant): Bringing new drugs to market is a lengthy and expensive endeavor. To increase the number of affordable new drugs available to consumers at reasonable costs, the efficiency of the drug development process must be greatly improved. We aim to develop a novel tool that addresses this challenge by rapid identification of drug targets and drug candidates for any given clinical phenotype utilizing our newly developed computational method. This method is based on a hypothesis that the tissue pattern of the expression of drug targets is essential for the unbiased matching of phenotypes to potential drugs. Matching complex molecular biomarkers of a disease to drugs thus depends on knowledge of the complete molecular signature (polypharmacologic fingerprint) of the drug or drug combination. Our approach is based on integration of large datasets from chemical and 'omics' databases and developing algorithms to link clinical phenotypes with drugs and drug-like chemicals. This is achieved by identifying the polypharmacologic profile of bioactive drugs and matching this drug-based profile to a phenotype-based tissue profile of a disease, which is derived from the computational analysis of 'omics' data. We have already developed and validated the 'drug' module of this system. Here we propose to (a) develop the 'phenotype' module, which will include the phenotype-tissue-target data via integration of clinical phenotypes, tissue-specific target expression, and GWAS and microarray data; (b) develop 'matching' algorithms for the modules that will build a direct link between drugs and clinical phenotypes as well as a ranking system for those matches; (c) validate this system using well-established clinical phenotypes and drugs for their treatment. When completed, the proposed research will result in a system that will allow users to instantly identify the ranked list of predicted drugs for any given clinical phenotype. Alternatively, it can be used to identify a list f phenotypes associated with any given drug/compound or drug/compound combination. This addresses a major challenge in the drug development process, improving the productivity and efficiency of the drug discovery phase, and it provides a useful and highly desired tool for the pharmaceutical industry and for academic research. Ultimately this technology will help in understanding the genomic contributions to the biology of disease and to accelerate the use of genomics to advance the science of medicine and the effectiveness of healthcare.