Detox from NIPS.
Getting back into more biological questions. Read two Reviews in Science. Van Oudenardeen on stochastic gene expression and Oates on patterning.
The first basically says that by looking at the histogram of mRNA distribution across many single cell measurements, we can infer something about the relative transition rates from Kon to Koff. Assuming a simple two state promoter model. Quite sensible.
The second outlines 4 type of pattern formation algorithms in developmental biology: gradient, diffusion reaction (Turing patterns), synchronization of oscillations, mechanical. Lots of experimental evidence for morphogen gradient; scant evidence for DR in embryos at least.
One interesting question would be to understand the computational powers of each of these natural “algorithms”.