Tuesday, January 27, 2009

Science Journalists and Junk DNA

The latest issue of SEED magazine concentrates on the idea that "Science Is Culture"—whatever that means.

One of the things it seems to mean is that good, accurate science reporting is not a high priority.

Junk DNA is one of those subjects that seem to bamboozle science journalists. They just can't seem to accept the possibility that much of our genome serves no purpose. One of the most extreme examples of this bias can be found in an article by Veronique Greenwood titled What We Lose.

The point of the article is that scientific models aren't perfect. They often over-simplify and, even more dangerous, they can exclude the very information required to refute the model. The example she uses is the software that will select what data to look at when the Large Hadron Collider (LHC) starts working. Greenwood's point is that the software might ignore the most interesting collisions because they aren't what scientists expect.

Here's how she explains the danger.
It wouldn't be the first time a standing model has excluded data that could revise it. In the 1980's, NASA analyses of the ozone layer flagged a great many data points as errors—values that seemed too low to be real, values that indicated a huge hole in the planet's protective layer. NASA scientists overlooked the possibility until an outside group published its discovery of the ozone hole in 1985. ....

Something similar happened in the 1990's when DNA that didn't code for proteins was labeled "junk." Noncoding DNA, biologist have since found, regulates protein-coding DNA.
THEME

Genomes & Junk DNA
No, Ms. Greenwood, that's not what happened. Junk DNA has been around for 35 years and it is well-established that much of our genome is composed of degenerate transposons and pseudogenes. There's good evidence that up to 90% of our genome may be junk, perhaps more.

Regulatory sequences have been known for over forty years. They cannot account for more than a small fraction of noncoding DNA. You are dead wrong when you claim that a function has been found for most junk DNA.





21 comments:

  1. OK, then -- Her argument or explanation may be flawed, but it doesn't follow that her conclusion (viz., that 'Junk' DNA can't be said to have no purpose at all), broadly, is therefore incorrect.

    "Shown to be false", and "not shown to be supported", are not equivalent logical states.

    Consider this: it was adaptively advantageous for primates in a fruit-rich forest to lose their ability, over generations, to synthesize ascorbic acid. The metabolic pathway was shut down. for inutility.

    But goats, horses, e.g., still make their own vitamin C. A diet of grass is not rich in vitamin C.

    The adaptive advantage is, why waste metabolic energy making something that is ubiquitous in your diet or readily available in your environment? -- And, vitamin C, for primates, is just one example amongst many metabolites for many families/genera/species, the synthesis of which has been abandoned, in the time since some or other ancestor shared with related organisms that still make that metabolite, whatever it is.

    And yet 'Junk DNA' has persisted in the eukaryotic genome across all genera for about a billion years, at great metabolic cost to all these possibly tens or even hundreds of millions of *types* of eukaryotic organisms.

    The metabolic advantage of ditching 'Junk DNA', would be tremendously adaptive -- a huge competitive advantage in a population that developed a mutation to rid itself of this implicitly wasteful metabolic load.

    We see this in every vat of engineered bacteria, where we have to add genes for antibiotic resistance to the plasmid with the gene for the desired metabolic product, or else the bacteria that spontaneously shed the plasmid will gain the upper hand, in the bacterial population in the vat.

    Yet this loss of 'Junk DNA' does not seem ever to have happened at any time amongst all the Eukaryotes.

    (Or, more precisely, if such a mutation has occurred, there does not sem to be any evidence that it has prospered).

    The necessary conclusion is that at least some 'Junk DNA' must have at least some purpose, even if it's as mundane as merely being scaffolding for the actual genome, or whatever.

    But there's more to it than that: the hubris implicit in "we don't know what it does, therefore it does nothing," is staggering.

    Scientists said that about, say, the human spleen and appendix, until the advances in immunology in the 1980s made it clear what those organs do. The fact scientist were ignorant of what they did, didn't mean they did nothing.

    I submit that in light of the metabolic load argument, an analogous situation applies to 'Junk DNA', today.

    It does something. We don't know what. Deal with it: find out what.

    A humble admission of ignorance, after all, lies at the root of all science. If we already "know" what it does (or doesn't) do, then how is our attitude different from that imputed to Galileo's intellectual opponents (Cremonini, etc.), who allegedly refused to look through the telescope supposedly because they already "knew" what they would see?

    There's a world of difference between "we don't know what 'Junk DNA' does", which is the scientifically appropriate statement, and, on the other hand, "we know 'Junk DNA' does nothing", which is probably not even a falsifiable statement, unless you wish to posit human omniscience, which would be a laughable claim.

    ReplyDelete
  2. "They just can't seem to accept the possibility that much or our genome serves no purpose."

    Ouch. And this coming from the guy who puts 'scientific accuracy' before all else. I'm surprised such a scientist would even use the term 'Junk DNA'.

    ReplyDelete
  3. Tim,

    But there's more to it than that: the hubris implicit in "we don't know what it does, therefore it does nothing," is staggering.

    This is a strawman. Do you think it appropriate to mischaracterise people's arguments in this way? It certainly seems to require a degree of hubris. Oh the irony.

    ReplyDelete
  4. Adaptively advantageous to lose the ability to synthesize Ascorbic Acid ?

    Really ?

    I would like to see a couple of number to support this. I doubt such a mutation could have been driven to fixation by selection.

    I think a strong counterargument to your claim can be made by simply looking at some general patterns of junk DNA. For example, large animals with very small effective population size tend to have a lot of junk in their DNA. Also, animals with energetically demanding lifestyle (example; birds, bats, ...) tend to have less junk in the DNA.

    It suggests that much of junk DNA can be explained by a reduction in the efficiency of natural selection, which is known to be pretty ineffective in small populations.

    In my opinion, most people won't accept junk DNA because they fail to understand how the mechanisms of evolution interact, everything is always about the simplest mechanism; selection. And many biologists also fail to understand this because, well, to be blunt, they just don't get the maths.

    ReplyDelete
  5. "Science is Culture" is their slogan.

    ReplyDelete
  6. @Tim

    Well you're relying on two things, first that maintaining the junk DNA is expensive. Now I don't know the relative energy costs but I guessing that DNA forms a very small percentage of our body mass, if the selective advantage is minute enough than there's no real significant advantage to ditching it.

    Second how do you get rid of it? Is their an attainable mechanism to drop all your junk DNA at once? If not would you have to get rid of it a gene at a time, if DNA is already a tiny amount of our metabolic expenditure how much smaller is the affect of a single gene? Furthermore, if we can't find a trick to make Junk DNA impossible even if we do get rid of a few bits we're also creating more at the same time.

    It general you want to be cautious before marking off something as useless. However, natural selection gives us a reason for why something is there. For the appendix either it has a selective advantage or had a selective advantage, that's the only reason for why we made it. But for Junk DNA if we have a good explanation for why it would show up even without a selective pressure then we don't need to assume a selective advantage to get it.

    Consider hearing a loud rumbling outside and you think it might be an accident. Then you notice a flash of lightning and another rumble. It doesn't mean that first rumble was thunder and not an accident, but it removes a lot of the evidence for that.

    Note, I haven't taken a biology course since high school so take my reasoning with a microscopic grain of salt.


    A question for Larry, how easy is it to categorize a bit of DNA? Is there a deterministic way to tell if a particular bit is coding/noncoding/regulatory?

    ReplyDelete
  7. "Consider this: it was adaptively advantageous for primates in a fruit-rich forest to lose their ability, over generations, to synthesize ascorbic acid. The metabolic pathway was shut down. for inutility."
    No, this is due to drift. It was broken, but had no effect on their survival. And leaving it in is no more harmful than any of the regions that codes for nothing because it's regulated. With high quantities of vitamin c it wouldn't even be making protein in the first place. Also, why was only a section the gene removed, but not the whole thing? If it was that detrimental then wouldn't the entire gene be removed instead of just a piece?


    "The metabolic advantage of ditching 'Junk DNA', would be tremendously adaptive -- a huge competitive advantage in a population that developed a mutation to rid itself of this implicitly wasteful metabolic load."
    Have you ever actually looked at the metabolic cost? During development DNA replication uses very little energy compared to other cellular processes, and during adulthood there is little cellular division making any extra cost insignificant. If there's no reason for it to disappear, selection cannot act upon it.

    "We see this in every vat of engineered bacteria,"
    No, just stop there, there is a reason bacterial genomes are small. It's because replication is a limiting step for reproduction, that is not the case with eukaryotes. There is an advantage for bacteria to have small genomes.


    "But there's more to it than that: the hubris implicit in "we don't know what it does, therefore it does nothing," is staggering."
    Considering that large amounts of the genome for mammals can be removed with no consequence I'd say it has little to no use
    link

    "Scientists said that about, say, the human spleen and appendix, until the advances in immunology in the 1980s made it clear what those organs do. The fact scientist were ignorant of what they did, didn't mean they did nothing."
    Source for the spleen claim please, as for the appendix claim, it is vestigial and there is no harm in it being removed. I have no issues with saying that a large amount of our DNA is vestigial and the result of duplication events (and subsequent mutations destroying the gene copies).

    "It does something. We don't know what. Deal with it: find out what."
    Considering the biggest tool for figuring out how genes work is by destroying them, and that it has been done with a large part of the genome, yeah, it's been looked into. Better yet, why don't you look into it?

    "A humble admission of ignorance, after all, lies at the root of all science. If we already "know" what it does (or doesn't) do, then how is our attitude different from that imputed to Galileo's intellectual opponents (Cremonini, etc.), who allegedly refused to look through the telescope supposedly because they already "knew" what they would see?"
    Comparing yourself to Galileo doesn't help your credibility, it just makes me think you're a crackpot. And unlike Galileo you have not given evidence of your claims. You can't accuse me of not looking at it when you supply nothing.

    ReplyDelete
  8. We see this in every vat of engineered bacteria, where we have to add genes for antibiotic resistance to the plasmid with the gene for the desired metabolic product, or else the bacteria that spontaneously shed the plasmid will gain the upper hand, in the bacterial population in the vat.

    Ditching a non-chromosomal element, like a plasmid, is mechanistically easier than discarding precisely this patch of chromosomal DNA. If you cut too much out of the chromosome, you can disrupt the neighbouring gene(s), which could be lethal. This plasmid analogy is incorrect when applied to loss of non-coding DNA.

    Second, this "energy cost of excess DNA" argument is getting too much weight, even though there is no evidence on the cellular scale that it is a meaningful cost. We know transcription and translation are also imperfect, leading to errors that need to be degraded (and a waste of energy). Under this cost logic, you would have expected the evolution of better and better systems. But we do not see that. The cell can "waste" this energy without significant survival costs.

    ReplyDelete
  9. When I said, "They just can't seem to accept the possibility that much of our genome serves no purpose."

    anonymous replied,

    Ouch. And this coming from the guy who puts 'scientific accuracy' before all else.

    In the name of scientific accuracy I demand that all possibilities be considered, not just that ones that conform to an adpaptationist bias.

    Do you have a problem with that?

    I'm surprised such a scientist would even use the term 'Junk DNA'.

    Perhaps you'd like to refute the scientific evidence in favor of junk DNA?

    You could start with the genetic load argument. That's always seems to be a good way to distinguish between people who know about this subject and people who don't.

    ReplyDelete
  10. Although Greenwood's enunciations about junk may be wrong, her point about models not being constructed to predict certain kinds of data and therefore lacking predictive power is very well-taken. For some eloquent and entertaining examples in science and finance, see Nassim Taleb's "The Black Swan"

    ReplyDelete
  11. A question for Larry, how easy is it to categorize a bit of DNA? Is there a deterministic way to tell if a particular bit is coding/noncoding/regulatory?

    Sorry, I'm not Larry so this won't be as good an answer. Also, I don't know what 'deterministic' means, but yes, there are reasonably good ways of telling whether a stretch of DNA is coding/noncoding/regulatory.

    Here, look at this (it helps if you like Gospel music):

    aaaaGocgcgcgcdowncgcgcgcMosesaaaaa

    As a very, very rough approximation, that's what DNA looks like. The letters (bases) are all strung together, but there are clues (usually conserved short stretches of DNA) to tell you where the 'words'(=genes) and 'punctuation'(=regulatory elements) should go. So you can parse a stretch of DNA to get a rough estimate of how much of each is present.

    And before anybody who makes a living doing this jumps on me for flaws in that analogy, I know, I know!

    ReplyDelete
  12. JHB:

    You are suggesting identifying regulatory DNA by comparison between sequences, relying on that if it's conserved (found in all/most the species examined), you argue it will have some function.

    While this is done, comparisons work better on more "obvious" things like gene coding regions.

    Most regulatory "elements" (words in your analogy) are rather short. Short enough that given the size of the genome, the small 'alphabet' (A, C, G, & T) and the frequency of mutations (changes in the sequence over time), it's hard, if not impossible, to distinguish the "real" ones from so-called "false positive matches": things that look the same, but do not function as a regulatory element.

    Imagine inserting random letters between words of several paragraphs that included several instances of the word 'an', then looking for 'an's: would you be able to tell if the 'an's you found came from the original sentences or the random junk added later? (I'm assuming both 'a' and 'n' occur reasonably often in the random letters added.)

    There are various ways of trying to get around this, but they all have limitations.

    Large-scale experimental methods can be used to detect what is functional. To give the gist of it, one approach is to randomly change or delete small pieces of DNA at a rate that you would expect only one change in each individual bacterium (etc) being tested, then test if the changes mattered. If the change results in a defect, you know it matters. If no defect is observed, it probably doesn't matter. (Obviously, this is limited by what the testing measures.)

    There are many different ways of doing this and these results are considered much more useful.

    ReplyDelete
  13. BioinfoTools:

    You're right. So am I.

    I made the hidden word analogy because that's the method which generates the '90% junk' claims.

    As you correctly say, it's better to do mutational analysis, but I'm not aware that that sort of high-detail analysis has been done in anything other than bacteria.

    As far as I know, the '90%' values come from scanning eukaryotic genomes with the sort of parsing software which I described and seeing how much of it doesn't match conserved gene or regulatory sequences; this becomes 'junk'.

    Sorry, you obviously know all this! I'm just making this comment for readers who don't.

    ReplyDelete
  14. JHB says,

    As far as I know, the '90%' values come from scanning eukaryotic genomes with the sort of parsing software which I described and seeing how much of it doesn't match conserved gene or regulatory sequences; this becomes 'junk'.

    If this is "as far as you know" then you are not qualified to debate this issue. I suggest you read my postings on Genomes and Junk DNA in order to get started on your long journey of education.

    The concept of junk DNA has been around for 35 years, long before large genomes were sequenced.

    ReplyDelete
  15. If this is "as far as you know" then you are not qualified to debate this issue.

    You need to work on that welcoming Socratic manner of yours.

    Still, BioinfoTools and you are more correct than I am - mouse, Drosophila (and human?) mutagenesis studies in the 1960s and 1970s - and I stand corrected on where the '90%' value comes from.

    ReplyDelete
  16. I've never understood the argument that since DNA replication takes 'so much energy' that it would be evolutonarily beneficial to have gottern rid of it. And lots of prety smart people have made such an argument. But I don't know why. Others ahve already addressed this, but to add, DNA replication really only happens when a cell undergoes mitosis. A typical cell, it seems to me, expends more energy on a daily basis just moving things across it's membrane than it does replicating it's 'junkDNA'. It is a pretty weak argument, regardless of who is making it.

    ReplyDelete
  17. JHB:

    You're right. So am I.

    I tried to politely point out you're not quite right. You can't have that both ways, I'm afraid.

    I made the hidden word analogy because that's the method which generates the '90% junk' claims.

    Whether you mean to or not, you have shifted the goal posts after the fact. In your earlier post, you presented your word analogy as being in reply to:

    Is there a deterministic way to tell if a particular bit is coding/noncoding/regulatory?

    I replied in this context, as I believe you did also; after all you quoted it as being what you were replying to... Furthermore, I specifically replied to identifying regulatory elements by sequence comparison (i.e. the last of the three listed in the question) as I did try to make clear.

    I believe you are confused, and I agree with Larry that you should read more before trying to make definitive statements.

    As you correctly say, it's better to do mutational analysis, but I'm not aware that that sort of high-detail analysis has been done in anything other than bacteria.

    (Large-scale, detailed, etc.) mutagenesis studies have been and are done on a wide range of species, including insects, fish, mice, plants, etc.

    A bit of advice: don't try present the other person's position. You've misrepresenting my position, even if it's out of lack of knowledge rather than malice. Again, I agree with Larry's suggestion that you read more first.

    ReplyDelete
  18. BioinfoTools said:

    You can't have that both ways, I'm afraid.

    Oh, enough already. I corrected myself and name-checked you three comments back, so lay off the concern trolling.

    ReplyDelete
  19. I'm not concern trolling, that term is derogatory and applies to a very different activity.

    Yes, I posted late. It happens. I'm working very long hours and my reply was written very late at night. I held off posting it until the morning to give myself a chance to re-read it, to avoid being too harsh on you. Ironic, eh? I make the effort to make sure I'm not being unfair and get labelled like that for my trouble! I have to admit this sort of thing is one of the reasons I rarely try help others on the WWW.

    ReplyDelete
  20. BioinfoTools:

    Leaving aside the etiquette of blog comments for a moment, please don't let my manners put you off posting here. Your molecular biology observations were right and are worth reading.

    Hope you have a good weekend and all the best with the sleep...

    ReplyDelete