Biologists should specialize, not hybridize

Why computational biologists should specialize in their dry skills and not learn other 'stuff', like wet biology, is excellently explained in this column by Assaf Zritsky in this issue of Nature.
As I use to say, computers are just tools like any other ones, such as microscopes or mass spectrometers. You don't ask your favorite microscope specialist to be also an expert in cloning or cell culture. Same thing with computers, we should collaborate not work in isolation.

Randomly distribution grants?

Randomly determining which grants are funded and which ones are not. This looks like a terrible idea, isn't it?
Well, may be not that bad, as suggested by J. Sills in this recent insight in Science. Indeed, currents systems have their limitations and randomly assigning money to the not so bad, but also not so good grants, the middle 60% might not be such a bad idee. Similar ideas have also been proposed to fix our 'democratic' system by example by D. van Reybrouck in his book 'Against the elections'. Why not the same in science?

Did mitochondria arrive late in eukaryogenesis?

Gabaldon just published an interesting phylogenetic analysis of eukaryotic signature proteins. The evolutionary signal that they could recover for those proteins does not support an early aquisition of mitochondria during eukaryogenesis, but to the opposite a late one, suggesting that most of the ancestral eukaryotic cell was already quite complex, when the mitochondria was aquired.