Conditional Random Fields for Predicting and Analyzing Histone Occupancy, Acetylation and Methylation Areas in DNA Sequences

Eukaryotic genomes are packaged by the wrapping of DNA around histone octamers to form nucleosomes. Nucleosome occupancies together with their acetylation and methylation are important modification factors on all nuclear processes involving DNA. There have been recently many studies of mapping these modifications in DNA sequences and of relationship between them and various genetic activities, such as transcription, DNA repair, and DNA remodeling. However, most of these studies are experimental approaches. In this paper, we introduce a computational approach to both predicting and analyzing nucleosome occupancy, acetylation, and methylation areas in DNA sequences. Our method employs conditional random fields (CRFs) to discriminate between DNA areas with high and low relative occupancy, acetylation, or methylation; and rank features of DNA sequences based on their weight in the CRFs model trained from the datasets of these DNA modifications. The results from our method on the yeast genome reveal genetic area preferences of nucleosome occupancy, acetylation, and methylation are consistent with previous studies.