The most recent junk DNA article describes a study by Gill Bejerano of Stanford University which I can’t find published anywhere (neither can ERV). Is Scientific American describing unpublished research on their website? Either way, they’re doing it poorly. The study described in the article identified non-coding sequences that were conserved across all sequenced mammalian genomes but missing from the chicken genome. Five percent of the sequences resemble known transposable elements, and the rest are of unknown origin. The Scientific American article said so much in nine paragraphs, and included the (unnecessary) word “junk” nine times (including twice in the headline and twice in a figure caption which is just a picture of a landfill).
Once again, I’ll say that I don’t mind the term “junk DNA” so long as it’s used appropriately. Throwing it around willy-nilly like every popular science writer seems to do is not the appropriate usage. Also, claiming that we have absolutely no freakin’ idea what non-coding DNA does is highly misleading. We know a fair bit about the function of non-protein coding DNA, including how it regulates transcription, how it encodes non-translated RNAs, and many studies have identified conserved non-coding elements. That said, a lot of it probably is junk. There are a ton of DNA sequences in any eukaryotic genome that serve absolutely no function, but they’re not deleterious enough to be purged from the population. So they’re just there, and that’s all there is to it.