Predicting and testing gene function in the human cell division cycle

描述（由申请人提供）：基因组学面临着将高通量方法产生的大量信息整合到内聚数据集以预测基因功能和途径的艰巨挑战。我们建议在未转化的人乳腺上皮细胞（HMEC）系中鉴定细胞周期调控基因，并直接将细胞周期调控基因与转化的HMEC系进行比较。我们计划使用细胞周期调节基因表达数据作为筛选，为未表征的细胞周期调节基因分配功能，然后通过细胞生物学方法实验测试这些计算生成的假设。为此，我们提出了在三种不同HMEC细胞系中鉴定细胞周期调控基因的实验，并在基于微扰的时间过程和芯片上的chlp实验中表征FOXM1转录因子控制的细胞周期调控电路。该研究的一个关键方面是进一步发展贝叶斯整合方法，以及新的计算工具，以分析基因组尺度的数据，探测细胞的生长和增殖，目标是为未表征基因提供准确的基因功能预测，以及对已知基因的更完整的描述。以功能预测为指导，将通过RNA干扰和高含量筛选对先前未表征的细胞周期调节基因进行分析，以测试功能预测。最后，每个未知基因的敲除将通过DMA微阵列杂交进行分析，以提供额外的“功能读出”表型。所有这些数据都将以迭代的方式整合回贝叶斯框架，以提高我们的预测能力。细胞周期控制是生长和发育的一个基本过程，它的放松会导致许多疾病，最明显的是癌症。因此，本提案所处理的问题对公共卫生很重要。确定受调控的基因及其生物学功能可能有助于开发抗癌疗法和生物标志物。

DESCRIPTION (provided by applicant): Genomics faces the daunting challenge of integrating vast amounts of information generated by high- throughput methods into cohesive datasets to predict gene function and pathways. We propose to identify the cell cycle-regulated genes in a non-transformed human mammary epithelial cell (HMEC) line and directly compare the cell cycle-regulated genes to those in transformed HMEC lines. We present a plan to use the cell cycle-regulated gene expression data as a screen to assign functions to uncharacterized cell cycle- regulated genes, and then to experimentally test these computationally generated hypotheses by cell biological methods. To this end, we propose experiments to identify cell cycle-regulated genes in a three different HMEC lines and characterize the cell cycle regulatory circuitry controlled by the FOXM1 transcription factor in perturbation-based time courses and ChlP-on-chip assays. A key aspect of the study is to further develop Bayesian integration methods, as well as new computational tools, to analyze genome- scale data probing cell growth and proliferation with the goal of providing accurate prediction of gene function for uncharacterized genes as well as a more complete description of the known genes. Using the functional predictions as a guide, previously uncharacterized cell cycle-regulated genes will be analyzed by RNA interference and High Content Screening to test the functional predictions. Finally, knockdown of each unknown will be analyzed by DMA microarrays hybridization to provide an additional "functional readout" for phenotype. All of this data will be integrated back into the Bayesian framework in an iterative manner to increase our predictive power. Cell cycle control is a fundamental process of growth and development and it's deregulation leads to many diseases, most notably, cancer. Therefore, the problems addressed in this proposal are important to public health. Identifying the genes that are regulated and their biological function may aid in the development of anti-cancer therapies and biomarkers.