Fig. 3

Comparison analysis of target genes identified by single normal distribution and mixture normal distribution. We utilized the NFE2 ChIP-seq data and gene expression data treated by NFE2 shRNA in K562 cells. a We compared the identified NFE2 target genes using three methods as follows: simple method, TIP with single normal distribution and TIP with mixture normal distribution. In the simple method, we identified target genes using a conventional peak-based method. The simple method identified 1597 target genes. We used various FDR thresholds (x axis) of TIP algorithm to select significant NFE2 target genes. The y axis represents the number of target genes. b We calculated the expression changes (absolute value of log ratios) of all genes for the cells treated with shRNA as compared to the untreated cells, and then we sorted the genes in decreasing order of absolute value of their log ratios. The genes on the left have greater absolute value of log ratios (WT vs. shRNA) and are therefore more responsive to NFE2 regulation. Given a threshold of absolute log ratio (x axis), the y axis shows the number of target genes satisfied the absolute log ratio threshold for each method