Carl Zimmer in his blog, The Loom:
The genetic study of human evolution really got off the ground in the 1980s. Allan Wilson of Berkeley and his colleagues compared the sequence of a gene in a sample of people and used their results to draw a genealogical tree. The gene came from mitochondria, energy-generating structures in our cells that also carry their own DNA. Wilson and his colleagues knew that we probably get all our mitochondria from our mothers. (Sperm apparently don’t deliver their mitochondria to eggs during fertilization—only 23 chromosomes that will end up in the nucleus of the fertilized egg.) If a woman’s mitochondrial DNA undergoes a mutation, she will pass that mutation down to her children, and her daughters will pass it down to their children. So finding people who share distinctive mutations allows scientists to see how closely related they are to one another. And since there was reason to think that the mutations arose at a relatively steady rate, they could even act as a molecular clock. If people shared an ancient ancestor, their mitochondria would be more different than if they shared a recent one.
The results of Wilson’s study were quite striking. The tree he and his colleagues drew showed that all of the genes on the deepest branches of the tree belonged to people of African descent.
More here.
