How does a cell know what kind of cell it should be?

It’s not what you’ve got, it’s how you use it. That seems to apply as much to cells as to people. As we grow from an embryo, lineages of replicating cells start from a versatile stem-cell state and end up self-assembling into tissues with specialised functions. This specialisation depends not on which genes the cells have, but on activating and deactivating the right genes at the right time. How does the machinery that turns genes into proteins know which part of the genome to read in any given cell type?

For gene regulation isn’t a matter of straightforward, digital transmission of information that relies on chains of lock-and-key recognition and binding of programmed biomolecules. Instead, it seems to involve something much more collective and dynamic at the molecular scale. Regulation is achieved not through selective molecular handshakes but from brief snatches of conversation within a bustling crowd of diverse molecules, gathered into a loose crowd amidst the genes themselves.

How that happens is still being debated. One view is that the molecular crowd is convened by the collective physics of phase transitions. Like the separation of oil from vinegar in salad dressing left to stand, phase separation in living cells leads to the appearance of tiny blobs of proteins and other molecules, which concentrates the components needed to switch genes on and off.