Genome-wide analysis of H3K36me3 and its regulations to cancer-related genes expression in human cell lines.

ABSTRACT

H3K36me3 is a histone modification known to mark active genes. To further understand the effects of H3K36me3 on gene expression levels, we develop predictive models to compute the correlation between the binding signal of H3K36me3 in each bin and the gene expression levels. We find that the bins with stronger H3K36me3 averaged-binding signals present higher correlative strengths with the expression levels of gene. And the higher correlative strengths appear in the downstream regions of the transcription start site. Moreover, we systematically compare the predictive abilities of 11 histone modifications to gene expression levels. The results show that H3K36me3 achieves a higher predictive ability than other modifications, and the higher predictive ability is robust across different mammalian cells and gene groups. Finally, in contrast to the two normal cell lines, our analysis finds that the predictive abilities of H3K36me3 are enhanced in 10 of the 13 bins for oncogenes and are decreased in 10 of 16 bins for tumor-suppressor genes in the cancer cell.