Since the last post was on how noise was good for some genes, it seems suitable to say that noise is, of course, bad for others:
Says that essential genes cluster in low-nucleosome areas due to the decrease in noise that such a cluster will cause. Basically, since misexpression of said genes will cause severe phenotypes, they are clustered in places where misexpression is less likely to happen randomly, even for a short time -- hence less noise. Kind of an interesting result. I like the idea that noise in gene expression is caused by chromatin remodelling. It's clever.