Figure 4

mRNA expression level correlates with the phosphorylation of RNAPII. (a) Expressions of genes assessed by expression microarray (left: GSM23372, middle left: GSM 161670, middle right: GSM246123) and RNAseq (right). Each column indicated classified genes by FPKM (= 0, or > 0) and RNAPII Ser2P and Ser5P binding state. Only those genes that had microarray expression data were analyzed; thus, there were 1,225/2,306 (53%) FPKM > 0/Ser2P(-)/Ser5P(-) genes, 348/588 (59%) FPKM > 0/Ser2P(+)/Ser5P(-) genes, 2,601/4,063 (64%) FPKM > 0/Ser2P(-)/Ser5P(+) genes, 4,727/6,505 (73%) FPKM > 0/Ser2P(+)/Ser5P(+) genes, 198/470 (42%) FPKM = 0/Ser2P(+)/Ser5P(-) genes, 232/667 (35%) FPKM = 0/Ser2P(-)/Ser5P(+) genes, 122/355 (34%) FPKM = 0/Ser2P(+)/Ser5P(+) genes, and 2,803 FPKM = 0/Ser2P(-)/Ser5P(-) genes available for analysis. Color key indicates gene expression value, yellow: over 90 percentile, black: median, blue: 10 percentile. The number of genes that were assessed was restricted by the microarray platform. (b) Histogram showing how many genes exist for each FPKM value. Y-axis indicates relative frequency in each category (RNAseq(+), Ser2P(+ or -), Ser5P(+ or -). Significant peak shift of the distribution of gene expression is shown according to each category. (c) Transcripts derived from genes that are categorized by Figure 3A were quantified by qPCR. The expressions of not only RNAseq(+) genes, but also some RNAseq(-) ChIPseq(+) genes are confirmed. (d) Assessment of the correlation between peak height of ChIPseq and FPKM value of RNAseq. Peak height for each gene was calculated by extracting the highest one that existed in the coding region. There is no obvious correlation.