At the NSF iPoLS meeting at Harvard, which JC, Sunny, and Curtis are presenting at, one of the best talks was given by Nigel Goldenfeld (UIUC).

The homochirality of certain biological molecules, especially amino acids in proteins (all L form) and sugars (all D form), has long been a mystery in biology. What processes in early evolution would have led to biological systems using only one form? Hypotheses have included autocatalytic mechanisms as well as even extraterrestrial origins. However, Goldenfeld has recently shown that there’s been a missing ingredient in many of these: the role of noise! Biological systems operate far from equilibrium, and we often find that noise (“stochasticity”) is a beneficial, even critical element, rather than a detrimental background to be minimized. Goldenfeld and colleagues showed that by incorporating two simple principles, self-replication and noise, homochirality can be established.

See a news story here and the original paper here.