No paper out yet, and not even any details made available, but this looks interesting:

Reduced genome works fine with 2000 chunks missing

To put a figure on how much of our DNA is non-essential, Vrijenhoek and his colleagues screened the genomes of 600 healthy students, searching for chunks of DNA at least 10,000 base pairs in length that were missing in some individuals. Across all the genomes, about 2000 such chunks were missing – amounting to about 0.12 per cent of the total genome.

Some people will over-interpret this as strong evidence for a majority of “junk DNA”. Comprising only 0.12% of the genome, it isn’t. However, as these are natural deletions >10kb, it gets around the objections to the deletion studies (i.e., that the conditions in the lab weren’t the same as the challenges faced in the wild). Then again, it may be that you can have one or two deletions and be ok because there is some redundancy, but if you were missing all of these bits you’d be in trouble. Others will dismiss it as an artifact or somehow not really testing the claim (read: dogmatic assumption) that all DNA is functional, but what else is new.