while variation exists:
choose the best variants in the population
probabilistically replace the worst variants with the best variants
The dominant metaphor in cellular biology is that proteins in macromolecular complexes are molecular machines that carry out work in the cell. It’s an excellent way to think about many cellular processes. For example, the proteins that generate ATP (chemical energy in the cell) from a proton gradient across cellular membranes have an uncanny resemblance to turbines generating electrical energy in hydroelectric dams.
Machines aren’t the only metaphor in the game. Richard Dawkin’s scientific masterpiece describes living things as vehicles for selfish pieces of data called replicators. There’s a lot of validity for this metaphor as well: the cell as a society of genes. Many genes within cells are involved in evolutionary arms races, especially viral relicts such as transposons. It’s odd to think of a machine where the different components are at war with each other. If one thinks of a cell as a society of genes, and organisms as a society of cells, one can imagine how a complex entity can survive and thrive even as components cooperate or compete with each other—although it’s odd to use a metaphor that implies that genes have agency, because they don’t.
I’m interested in how a ‘society of genes’ perspective might help one understand why and how cells are built the way they are. But this perspective won’t necessarily add much to the machine picture of cells. The reason why is that when the reproduction of all genes is completely linked, all genes have to contribute to the whole in order to reproduce their own information. Over long periods of time, useless genes face the eventual fate of loss or deletion. On the other hand, if there’s a lot of horizontal gene transfer (or gene duplication followed by recombination), some genes could successfully spread without contributing to the function of the whole. Lytic viruses are the ultimate example of selfish agents that hijack the normal functions of a cell in order to replicate themselves. It’s also interesting to note that there are many cases in which cells have domesticated genes that were previously viruses or transposons—many of the components in cellular machinery and information processing pathways have been co-opted from such antagonistic sources.
Evelyn Fox Keller has written an excellent book, ‘Making Sense of Life,’ that describes how the definition of life has itself changed over the past 300 years of scientific inquiry, and how the metaphors and concepts that people use to understand life have shifted as society changes. For example, it would be completely weird and alien for a person in 1000AD to think of life as being made up of teeny tiny cellular machines.
The metaphors that scientists keep in the back of their heads are important because they guide the kind of questions that scientists ask. In general I think it makes sense for biologists to use the machine metaphor, but I do think that it has one big drawback: it makes systems appear more controllable and predictable than they truly are. It’s intuitive to imagine engineering life when one thinks of biology as a medium for building machines. But when one thinks about life as a complex system more akin to an economy or society, it's natural to be somewhat skeptical about biological machines.