ANOVA-Based Approaches to Time-Series Microarray Data

描述（由申请人提供）：在考虑生物医学研究中转录反应的动态变化时，需要将统计显著性与生物学相关性结合起来。为此，将开发基于方差分析（ANOVA）的方法，通过集中对比和治疗条件内部和之间的趋势测试，在时间过程微阵列数据中检测差异表达的基因。这些基于方差分析的效率将与经验贝叶斯方法和b样条拟合分析进行比较，以综合各种方法的最佳特征的新方法。接下来，方差分析方法得出的基因列表与已知生物过程之间的统计学相关性将用于揭示果蝇和小鼠脑组织样本中睡眠剥夺研究中的重要关系，以及UAMS-33小鼠细胞培养中影响骨细胞和脂肪细胞分化的因素。最后，将设计验证性实时定量PCR方法，并将其应用于UAMS-33小鼠细胞系实验中，采用更高频率的时间点采样，以建立精细表征该系统时间动态的图形模型。

DESCRIPTION (provided by applicant): There is a need to integrate statistical significance with biological relevance when considering dynamic changes in transcriptional response in biomedical research. Towards this end, methodologies based on analysis of variance (ANOVA) will be developed to detect differentially expressed genes in time course microarray data through focused contrasts and trend tests within and between treatment conditions. The efficiency of these ANOVA-based analyses will be compared with empirical Bayes methods and B-spline fitting analyses to synthesize new methodologies that incorporate the best features of the various approaches. Next, statistically relevant associations between gene lists derived from the ANOVA method and known biological processes will be used to uncover relationships important in studies of sleep deprivation in brain tissue sampled for flies and mice and factors influencing osteocyte and adipocyte differentiation in UAMS-33 mouse cell cultures. Finally, confirmatory real-time quantitative PCR assays will be designed and applied to the UAMS-33 mouse cell line experiments with higher frequency time point sampling in order to build graphical models that finely characterize the time dynamics of this system.