根据周期表达基因的周期性和峰值特点,提出了一种将microarray时序表达数据划分为若干个基因表达周期,并对周期内的峰值特点进行评估以识别周期表达基因的方法,能有效减小microarray实验时的噪声干扰。选取了三组广泛使用的时序表达数据和一组可靠的周期表达基因集合对该方法的效果进行了测试,并与三种典型的周期表达基因识别方法的效果进行了比较。该方法能有效地从各种microarray时序表达数据中识别周期表达基因。
According to the periodic peak trait of periodically expressed genes, we have developed a novel method which can deal with the disturbance in microarray time series data, and obtain more accurate identification of periodically expressed genes. A comparison between our method and other three methods was made by applying three different cell cycle expression data using one benchmark. The results demonstrate that our method is effective in the analysis of expression data with disturbances.