Debunking the idea that most humans only use 10 percent of our brains, which is a starting point for the Scarlett Johansson/Luc Besson movie Lucy, was popular last week.
Many media outlets and popular Web sites took on this task. Emoryâ€™s Krish Sathian â€“ known for his work on rehabilitation, how the brain processes sensory experiences and how we understand metaphors â€“ does an able job of it in the video below.
But a related question is still a matter of debate: how much of our DNA do we â€œuseâ€? This is an important question for geneticists because it seeks to define the most productive mutation hunting grounds.
A study published in PLOS Genetics last week concluded that just 8.2 percent of the human genome is constrained during evolution and is likely to be â€œfunctionalâ€. TheÂ press releaseÂ on this paper pointed out sharply that this contrasts with the more expansive analysis from the multinational ENCODE project, which assigned some biochemical function to 80 percent of the human genome.
Three points to keep in mind:
*A large portion of the human genome is taken up by repetitive elements that are, essentially, spam. (When you delete a spam email, are you using it? But even spam can jog your memory.)
*The spectrum of DNA that is important varies according to tissue and cell type.
*Thereâ€™s a difference between how much DNA is somehow â€œactiveâ€ (lots) and how much is sensitive to mutations (less).
Only a small fraction of the human genome â€“ less than 2 percent — codes for proteins. Proteins are the machines that drive the cellâ€™s functions, and this is the part of the genome where we can expect mutations to have significant effects. Beyond the protein-coding regions are control buttons like promoters and enhancers, also likely to be important functionally. Non-coding but regulatory RNAs also occupy this in-between zone.
In the outer zone, most of the genome is transcribed â€“ meaning that an RNA polymerase enzyme factory plows through and copies the DNA into RNA, enough for that RNA to be detected.
This PNAS review from ENCODE authors has some nice diagrams that make these distinctions clear. Thanks to Michael Zwick and Chris Gunter for discussing this question with me.