NEW STRATEGIES FOR DRUG DISCOVERY: GENE EXPRESSION PROFILING AND BIOINFORMATICS

这个项目的目的是整合三种类型的癌症分子药理学信息：(1)NCIS药物发现计划中测试的化合物的癌细胞抑制活性模式；(2)测试化合物的分子结构特征；(3)用于测试的细胞中可能的分子靶标和活性调节剂--通过基因和蛋白质表达谱进行评估。这些结果对药物的发现和设计、可能的个体化治疗以及对癌细胞分子标志物的临床研究具有重要意义。到目前为止的分析结果包括：(I)能够根据DTP-60细胞系癌症药物筛选的活性模式预测药物作用机制的神经网络；(Ii)集成了NCI测试的化合物的化学结构、活性和可能的分子靶标信息的程序包(DISCOVERY)。顾名思义，DISCOVERY的设计目的是在迄今测试的60,000种化合物中寻找新的作用机制；(3)神经网络和统计方法，以预测II期药物的临床活性(应该强调的是，仅具有中等的灵敏度和特异度)；(4)根据筛选中的活性模式，可评估药物的临床活性；(4)用于流畅搜索关于基因、药物和疾病的文献的程序包MedMiner，以帮助进行表达谱研究；(5)在NCIS DIS数据库中进行定量构效关系研究和药效团搜索，以寻找HIV-1整合酶和人类拓扑异构酶的新抑制剂；(Vi)与癌症基因组解剖学项目中的临床肿瘤标本建立接口。迄今的实验结果包括(I)为NCI药物发现计划中的60个细胞系开发一个大型蛋白质表达数据库(通过2D-Gel电泳法，与大型生物公司Leigh Anderson博士合作)；(Ii)使用cDNA微阵列(与斯坦福大学和Synteni公司的合作者)为60个细胞系建立一个10,000个基因的信使核糖核酸表达数据库；(Iii)使用Affymetrix寡核苷酸芯片(与怀特黑德基因组中心的合作者)建立一个42,000个基因的信使核糖核酸表达数据库；(4)增加关于其他细胞类型实验数据的中央数据库，包括内皮细胞、多药耐药癌细胞、p53等基因细胞组，以及用于艾滋病毒药物发现的细胞类型；(5)开发基于质谱学的方法，以确定和表征凝胶中亚皮克摩尔量的蛋白质。作为一个整体，该项目代表了与NCI发展治疗计划以及其他一些实验室的合作。Http://discover.nci.nih.gov.上的工具和数据-艾滋病、癌症、药物发现、凝胶电泳、基因表达、遗传算法、神经网络、蛋白质表达、微阵列、

The aim of this project is to integrate three types of information on the molecular pharmacology of cancer: (1) patterns of cancer cell inhibitory activity for compounds tested in the NCIs drug discovery program; (2) molecular structural features of the tested compounds; (3) possible molecular targets and modulators of activity in the cells used for testing -- as assessed through gene and protein expression profiling. The results have implications for drug discovery and design, for possible individualization of therapy, and for clinical studies of molecular markers on cancer cells. Analytical results to date have included the following: (i) Neural networks able to predict mechanism of drug action on the basis of patterns of activity in DTPs 60-cell line cancer drug screen; (ii) A program package (DISCOVERY) that integrates information on the chemical structure, activity, and possible molecular targets of compounds tested by NCI. As the name suggests, DISCOVERY was designed to search for novel mechanisms of action among the 60,000 compounds tested to date; (iii) Neural nets and statistical methods to predict (with only moderate sensitivity and specificity, it should be stressed) the clinical activity of phase II- evaluable drugs on the basis of patterns of activity in the screen; (iv) A program package MedMiner for fluent searching of the literature on genes, drugs, and diseases to aid in expression profiling studies; (v) QSAR studies and pharmacophore searches in the NCIs DIS database for new inhibitors of HIV-1 integrase and human topoisomerase ; (vi) creation of an interface with clinical tumor specimens in the Cancer Genome Anatomy Project.Experimental results to date have included (i) Development of a large protein expression database for the 60 cell lines in the NCI drug discovery program (by 2D-gel electrophoresis, in collaboration with Dr. Leigh Anderson, Large Scale Biology Corp.); (ii) generation of a 10,000-gene mRNA expression database for the 60 cell lines using cDNA microarrays (with collaborators at Stanford University and Synteni, Inc.); (iii) generation of a 42,000-gene mRNA expression database using Affymetrix oligonucleotide chips (with collaborators at the Whitehead Genome Center); (iv) Addition to that central database of experimental data on other cell types, including endothelial cells, multi-drug resistant cancer cells, p53-isogenic cell sets, and cell types used for HIV drug discovery; (v) Development of methods based on mass spectrometry to identify and characterize proteins in the gels in sub-picomole amounts. This project as a whole represents a collaboration with the NCI Developmental Therapeutics Program, as well as a number of other laboratories. Tools and data at http://discover.nci.nih.gov. - AIDS, Cancer, Drug Discovery, gel electrophoresis, gene expression, genetic algorithm, neural network, Protein expression, microarray,