From the paper:
> Analysis of two high resolution nucleosome maps revealed
> strong signals that—even though they do not constitute a
> definite proof—are at least consistent with such a view.
Also, the premise of this paper is clearly something only a physicist would think was interesting. And I'm not the least bit surprised this wound up in PLoS.
This just in: Modeling DNA using what we already knew about DNA confirms what we already knew about DNA.
-----
Edit: OK, I've gone back and actually read the whole paper. I can't say I'm particularly swayed by this paper and I'm not entirely sure the point. But, in a pattern I'm seeing throughout biology, a non-biologist modeler just realized that bio is cool and spent the weekend modeling something. Overall, I'm not sure what this brings to the table that merits being published?
Here's my reading of the main points:
1. Some DNA sequences are bendier than others [Not a finding of the paper, already known]
2a. There are histone binding sites and "promoters" for nucleosome formation [already known]
2b. Through modeling, Eslami-Mossallam et al., claim a major factor in nucleosome binding is whether the DNA can (thermodynamically) easily bend around the histone complex. OK, fine, the models are maybe more specific but this preference (and the motifs that comprise the hotspots) were more or less known.
3. DNA sequence determines the physical and genetic properties... Yes, we know.
4. The degeneracy of the genetic code allows it to encode additional information. We already knew this, in fact, we know the mechanical properties of RNA encode functional roles (See, programmed ribosomal frameshifts and the DNApol gmaa/tau subunits)
5. [The real shocker] DNA in real life organisms encodes both genes and its own structure.