December 10, 2007


Study finds humans still evolving, and quickly (Karen Kaplan, 12/10/07, Los Angeles Times)

By examining more than 3 million variants of DNA in 269 people, researchers identified about 1,800 genes that have been widely adopted in relatively recent times because they offer some evolutionary benefit.

Until recently, anthropologists believed that evolutionary pressures on humans eased after the transition to a more stable agrarian lifestyle. But in the last few years, they realized the opposite was true -- diseases swept through societies in which large groups lived in close quarters for a long period.

Altogether, the recent genetic changes account for 7% of the human genome, according to the study published in the Proceedings of the National Academy of Sciences.

The advantage of all but about 100 of these genes remains a mystery, said University of Wisconsin-Madison anthropologist John Hawks, who led the study.

There just has to be an advantage in here somewhere....

Posted by Orrin Judd at December 10, 2007 8:41 PM

How can 7% of the human genome be recent when we have been told repeatedly that humans and chimpanzees have 96-99% (depending on who is making the claim) of their DNA in common? Something does not compute.

Posted by: jd watson at December 10, 2007 9:10 PM

Leaving aside the probability that these scientists are on crack, one reasonable explanation could easily found by extending Julian Jaynes hypothesis that language preceded "consciousness."

Once "conscious," man would direct his own evolution much more effectively - first somewhat unconsciously (not realizing his heightened consciousness was leading him) and then consciously (stripping out bad cells, diseases, abortion, etc.)

If you want to see where it all leads, rent "Idiocracy" - one of the most accidentally conservative comedies you'll ever see.

Posted by: Bruno at December 10, 2007 9:53 PM

I didn't see the 7% figure mentioned anywhere in the original paper so I don't know where that comes from.

Until a lot more studies are run which corroborate the results and work with much bigger sample sizes, it's not possible to say more other than it's potentially interesting.

There are some major questions about the statistical methodology they used, does look like they structured it to give the results they wanted.

Posted by: Ali Choudhury at December 11, 2007 6:21 AM