I’m not convinced that New Scientist gets it yet


Finally, they’ve come right out and said what we knew all along: most of our DNA has to be junk. I guess that’s progress, but they’re not doing a good job of explaining it.

After 20 years of biologists arguing that most of the human genome must have some kind of function, the study calculated that in fact the vast majority of our DNA has to be useless. It came to this conclusion by calculating that, because of the way evolution works, we’d each have to have a million children, and almost all of them would need to die, if most of our DNA had a purpose.

None of the biologists I know have been arguing for ubiquitous functionality, but I know they’re out there, so that’s kind of a strange opening: it’s as if the only way they know how to frame the story is as some kind of real conflict (see also every NS article about evolution vs. creationism). I don’t know where the 20 year timing comes from, either. JBS Haldane died 53 years ago, and he worked out this argument long before his death.

But worst of all, they just plop out this claim that we’d “each have to have a million children, and almost all of them would need to die, if most of our DNA had a purpose”. OK. Reading this as a naive layman, WHY? They present the conclusion with none of the evidence or logic behind it; there is no explanation here. The key part of the story that Dan Graur explained is that we know the mutation rate of human genes, and we can calculate the cost to the population of carrying around suboptimal genes, and we can estimate how many children you’d have to have to compensate for that load of mutations, and the load is going to depend on how many genes are present. It’s easy to put an upper bound on the number of genes we have, given our mutation rate and how many children an individual can have (hint: there’s no way you can have a million kids.)

The logic is clear and convincing, but you have to present it if you’re trying to communicate the science.

I feel like I’m grading an exam. Yes, you got the correct answer, but I’m not convinced that you understand how you arrived at it, and aren’t just regurgitating something you memorized.

Comments

  1. says

    New Scientist’s “Darwin was wrong” cover from 2009 convinced me they’re more interested in sensationalism than science. I’m still disgusted by it.

  2. amstrad says

    Okay. So, super ultra genetics layman here. Apologies for possible inappropriate use of genetic terminology, but here is how I think of “junk DNA”:

    DNA has no purpose, other than to function as a convenient package for replication and transcription of genes. Genes have no purpose other than some genes code for proteins that have phenotypes with real and tangible effects on whether that particular gene makes it to the next generation. Other segments of DNA get a free ride due to crossing over and the locality with previously mentioned genes. The free ride DNA may even code for proteins. These superfluous proteins may not be “useful” or active currently in bio pathway, but it may become so in the future. All these extra “unnecessary” proteins may be useful as building blocks of future bio pathways and emerging phenotypes. The DNA replication process has not stripped out these extra bits, because 1) it’s impossible to tell at replication which parts are what and 2) DNA that could strip out “junk DNA” might be less adaptive to change.

  3. says

    Their “Darwin was wrong” cover from 2009 convinced me that New Scientist is more interested in sensationalism than science. I’m still disgusted by it.

  4. richardemmanuel says

    If it were all functional, one would have to wonder what all the functionals of largerthanhumangenome creatures were. I’m keeping an eye on the locusts.

  5. says

    But it’s not junk! Friendly aliens encoded the Encycolpædia Galactica (all eight hundred thousand volumes of it) in our DNA to be available once we developed enough to read it. Mutation is actually periodic updates.

  6. methuseus says

    (hint: there’s no way you can have a million kids.)

    This seems like either a challenge or a really weird math problem. I don’t feel like figuring out if it’s possible for a man to physically father a million kids. For a woman there’s the obvious issues of gestational time.

  7. robert79 says

    You would have to impregnate about 30 women a day for the next 100 years. It might be feasible if you lived really long, started really young, and make daily visits to a sperm bank.

    I think sperm banks have rules about this though, too many donations from a single donor increases the chance of (half) sibling match ups a generation later.

  8. methuseus says

    You would have to impregnate about 30 women a day for the next 100 years. It might be feasible if you lived really long, started really young, and make daily visits to a sperm bank.

    The biggest problem with that is that I don’t believe a man produces enough sperm each day to impregnate 30 women… So basically PZ is right and it’s a physical impossibility. The highest realistic number for an especially virile man is likely one woman per day. That would result in 365 * number of years impregnating women. Likely a quarter million would be the highest possible. But that would take a lot of logistics including all the women being amenable to having children with this man.

  9. mijobagi says

    What is the accepted percentage of junk DNA? I actually thought it was higher than 75%!

  10. Kristian Ekeroth says

    I feel I still have to ask though…

    “we can calculate the cost to the population of carrying around suboptimal genes” Ok, but what does that say about non-coding DNA that still may have a function (“Function” defined broadly)? One possible function for non-coding parts could be to adjust the position of the coding parts in the DNA molecule. It does matter what parts are accessible to transcription, and when, and what other parts are also likely to be transcribed at the same time. Having large chunks of “spacer” DNA might not cost that much, so for most organisms the risks of cutting it out may not be worth it. (Some viruses have extremely effective genomes though, with reading frames in both directions, because it is advantageous to fit into tiny capsids and so on). In many cases more effective “solutions” would be possible, but evolution builds on whatever works. So genome sizes could differ a lot even if large amounts of DNA had such a function, through accumulation of more or less important “adjustments” over time.

    Of course some of it will just be “real” junk making no difference at all. And the effects of removing some “spacer” DNA may be small. But if we do count spacing as a non-junk function, is it really possible for us today to say how much of DNA is “real junk”?

  11. emergence says

    Lest any creationists get it into their heads that the ENCODE results were “censored” to protect evolutionary biology, I think it’s important to emphasize that this is an issue of basic population genetics and is a problem even within the limited microevolution they acknowledge the existence of. Genetic load doesn’t stop being an issue if you place arbitrary limits on how much evolution can change an organism.