Chen Hong-Chu, Lee Hsiao-Ching, Lin Tsai-Yun, Li Wen-Hsiung, Chen Bor-Sen
Department of Electrical Engineering, National Tsing Hua University, Hsinchu, Taiwan.
Bioinformatics. 2004 Aug 12;20(12):1914-27. doi: 10.1093/bioinformatics/bth178. Epub 2004 Mar 25.
Genome-wide gene expression programs have been monitored and analyzed in the yeast Saccharomyces cerevisiae, but how cells regulate global gene expression programs in response to environmental changes is still far from being understood. We present a systematic approach to quantitatively characterize the transcriptional regulatory network of the yeast cell cycle. For the interpretative purpose, 20 target genes were selected because their expression patterns fluctuated in a periodic manner concurrent with the cell cycle and peaked at different phases. In addition to the most significant five possible regulators of each specific target gene, the expression pattern of each target gene affected by synergy of the regulators during the cell cycle was characterized. Our first step includes modeling the dynamics of gene expression and extracting the transcription rate from a time-course microarray data. The second step embraces finding the regulators that possess a high correlation with the transcription rate of the target gene, and quantifying the regulatory abilities of the identified regulators.
Our network discerns not only the role of the activator or repressor for each specific regulator, but also the regulatory ability of the regulator to the transcription rate of the target gene. The highly coordinated regulatory network has identified a group of significant regulators responsible for the gene expression program through the cell cycle progress. This approach may be useful for computing the regulatory ability of the transcriptional regulatory networks in more diverse conditions and in more complex eukaryotes.
Matlab code and test data are available at http://www.ee.nthu.edu.tw/~bschen/quantitative/regulatory_network.htm
全基因组基因表达程序已在酿酒酵母中得到监测和分析，但细胞如何响应环境变化来调控全局基因表达程序仍远未被理解。我们提出了一种系统方法来定量表征酵母细胞周期的转录调控网络。出于解释目的，选择了20个靶基因，因为它们的表达模式随细胞周期呈周期性波动，并在不同阶段达到峰值。除了每个特定靶基因最显著的五个可能调控因子外，还表征了在细胞周期中受调控因子协同作用影响的每个靶基因的表达模式。我们的第一步包括对基因表达动力学进行建模，并从时间进程微阵列数据中提取转录速率。第二步包括找到与靶基因转录速率具有高度相关性的调控因子，并量化已鉴定调控因子的调控能力。
我们的网络不仅识别出每个特定调控因子作为激活剂或抑制剂的作用，还识别出调控因子对靶基因转录速率的调控能力。高度协调的调控网络已识别出一组负责细胞周期进程中基因表达程序的重要调控因子。这种方法可能有助于在更多样化的条件和更复杂的真核生物中计算转录调控网络的调控能力。
Matlab代码和测试数据可在http://www.ee.nthu.edu.tw/~bschen/quantitative/regulatory_network.htm获取
