Evolutionary biologists use so-called molecular clocks to determine when species branched off from family trees. These evolutionary timepieces, based in part on the rate at which DNA mutates, are far less accurate than atomic clocks, however, causing some scientists to question the value of the dating technique (Science NOW 28 November 2001 ). Now, two evolutionary biologists say they've found a better way to estimate DNA mutation rates, at least for mammals, a result that should give new precision to evolutionary studies.
To determine the rate at which DNA mutates, researchers compare different versions of the same gene from two to several dozen species. They count genetic differences and then divide by the best guesstimate of how long ago the species diverged, calibrated in part on the fossil record. But genes mutate at different rates, since, for example, some can withstand mutations better than others. As a result, mutation rates can vary between species by 10-fold.
Dissatisfied with such wobble in the basic gears of the molecular clock, evolutionary biologists Sudhir Kumar and Sankar Subramanian decided to establish a more reliable mammalian mutation rate based on as many gene sequences as possible. They started with more than 17,200 DNA sequences, representing 5669 genes from 326 mammalian species. Some 60% were disqualified based on certain criteria--if, for example, a gene in mice was in a more mutation-prone chromosomal region than the same gene in humans, the researchers decided that gene would add too much noise to the analysis. The remaining sequences yielded a mutation rate that equaled one of the existing estimates--2.2 changes in a DNA sequence a billion bases long, per year. This rate varied by less than 10% between mammalian species, they report 15 January in the online edition of the Proceedings of the National Academy of Sciences.
Other researchers were somewhat skeptical. Cutting out so many genes from the analysis and then saying the rate applies to all mammalian genes is a little unfair, says evolutionary biologist Michael Lynch of Indiana University in Bloomington. He says the results will be useful for timing some events in the evolution of mammals, however. Clearly, biologists aren't ready to give molecular clocks up--the clocks may not be Swiss, but they take a licking and keep on ticking.
Kumar's lab