A Statistical Thermodynamic Approach for Predicting The Sequence-Dependent Nucleosome Positioning along Genomes

The distribution of nucleosomes, the fundamental repeating units of chromatin, along the eukaryotic genome rules replication, transcription, repair, and regulation processes through modulation of DNA accessibility. Genome-wide maps provide much information about the factors that direct nucleosome positioning. However, the experimental nucleosome maps do not permit to conclude unambiguously that the DNA sequence of the eukaryotic genomes encodes nucleosome positioning and organization. A possible way to disclose this important issue is to develop theoretical models capable of predicting nucleosome positioning in terms of the DNA sequence. Toward this goal, we propose a physical model for predicting nucleosome thermodynamic stability in terms of DNA sequence. The model, based on a statistical mechanical approach, allows the calculation of the canonical ensemble free energy involved in the formation of each nucleosome along a DNA tract. The theoretical nucleosome distribution along genomes was compared with the experimental positioning maps of yeast genome. The results are comparable with those obtained with pure statistical models based on identifying some recurrent sequence features obtained from the statistical analysis of a very large pool of nucleosomal DNA sequences. However, our model based on the physical properties of the DNA such as curvature and flexibility appears universal and applicable to any genomes without rearrangements.