The RNA world hypothesis has been around for quite a long time as an origin of life theory (alongside others such as the iron-sulfur world and the newer aromatic world theories). Actually, the concept dates back as far as 1963. The idea is, essentially, that RNA can act as an enzyme, store information, and self-replicate. As such, many have favoured the idea that strands of RNA might have been evolving chemically long before cellular life developed.

Well, a research group at the Scripps Research Institute have hacked some RNA, creating a kind of molecular 'ecosystem'. It's not actually a living system, but a collection of molecules which evolve and compete the way living organisms do.

The team have been tinkering with these molecules for the past 8 years, creating pairs which require each others' help to reproduce. Effectively, sexual reproduction in molecular form. As a result, once these molecules were in solution, some of them gave rise to random mutations. Just as in living ecosystems, most mutations rapidly died out, but a select few proved to be beneficial. So beneficial, in fact, that after a mere 77 replications all the original molecules were extinct. A variety of forms came into existence with 3 highly successful molecules dominating the population.

Chemistry acting like biology. Fascinating. Simply fascinating.

The big question though is, could any other molecules could act this way? RNA is, after all, fairly advanced as molecules go. At what stage in a prebiotic system does chemistry start to act in a lifelike way?

The journal paper was published recently in Science, if you happen to have a subscription (or a nearby library).
Tip of the hat to ranka for posting that link!