Identifying the Significant Change of Gene Expression in Genomic Series Data for Epistasis Peaks

The changes in gene expression under microarray technology are valuable to recognize and study the evolution process of species or diseases. Among the existing methods of analyzing the changes in gene expression, statistical method is one of the common and accurate approaches. This paper presents a step-by-step protocol to use Biopeak, a statistical tool to identify and visualize any significant impulse-like change of the gene expression in genomic series data. Biopeak focuses on the temporal features of the gene expression as signals. Through the statistical approaches including finding the local maximum and subsequent filtering, the potential changes are detected as the peak of signals. To filter the outliers and mark the significant changes, the correlation heatmap and clustering approach can be applied. Biopeak also provides several clustering techniques with different cluster abilities for result comparison. The step-by-step application of Biopeak is carried out by running the dataset of human epithelial cells in response to heat. The results of the peak detection, the correlation heatmap, and the clustering are demonstrated.