Identification of peculiar and common effects of histone modifications on transcription.

Histone modifications (HMs) play an important role in controlling eukaryotic gene expression and next generation sequencing (NGS) has greatly advanced the research on this topic with generating many high-resolution maps for HMs. Here, we use these maps to analyze the relationship between HMs and transcription. By incorporating various segments of genes into analysis without restricting the scope only in the promoter region, we have collected more comprehensive data and captured some details of this process. A position effect of gene regions has been revealed and it can even inverse the property of some HMs from activating to repressing genes such as the cases of H3K4me3, H3K36me3 and H3K14ac. Especially H3K36me3, its dual character on gene transcription makes it able to serve as a criterion to distinguish high and low expressed genes. We also study the general property of different HMs based on the comprehensive data. Using exploratory factor analysis (EFA), we have extracted 4 latent structures underlying the HMs, which are able to represent their activating and repressing effects concisely. These 4 factors have fine properties in the aspects of distinguishing high and low expressed genes, predicting transcription level and identifying genes with unique attributes such as stable RNA generating genes found to have a close relationship with lifespan of organisms here. In summary, while the position effect associated HM peculiarities demonstrate some details of the complex HM regulation network divergently, the common factors catch the nature of the network convergently. This deepens our understanding on the HM-transcription relationship.