Predicting Nucleosome Position

A group including Jonathan Widom at Northwestern University and Eran Segal of the Weizman Institute used sophisticated computer modeling along with experimental approaches to uncover DNA-encoded rules guiding nucleosome position. ¹ The verdict: several dinucleotide sequences, because of their inherent bendability, can be used to predict roughly 50% of nucleotide positions. The work, may lead to better prediction of nucleosome binding and may influence such things as transgene design. Bradley Bernstein, a Faculty of 1000 member and assistant professor at Massachusetts General Hospital calls it an "impressive analysis." He adds "The paper identified interesting correlations that are likely to reflect causal relationships."

"It suggests that the positioning of nucleosomes is in some cases deliberate and functional. So you regulate accessibility to DNA in a very precise way and you can then control whether transcription factors have access and whether a gene can be activated and so on?. I think the interesting thing about this paper is that, to a surprising extent, DNA sequence does appear to be important. That being said it's clearly only a piece of the equation?. The role of the sequence does appear to be modest according to the correlations in the paper. Nonetheless, the effects will need to be considered in the development of increasingly sophisticated models of gene regulation."