In tomorrow’s New York Times, I have an article about how to reconstruct a genome that’s been gone for 80 million years. The genome in question belongs to the common ancestor of humans and many other mammals (fancy name: Boreoeutheria). In a paper in this month’s Genome Research, scientists compared the same chunk of DNA in 19 species of mammals. (The chunk is 1.1 million base pairs long and includes ten genes and a lot of junk.) The researchers could work their way backwards to the ancestral genetic chunk, and then showed they could be 98.5% certain of the accuracy of the reconstruction.

There are some pretty astonishing implications of this work. For one thing, it should be possible