There is a painful assumption of progress in many interpretations of evolution — and sometimes it’s by people who ought to know better. T. Ryan Gregory finds a ghastly example of a figure that, by cherry-picking the data and doing a little suggestive ordering of the presentation, makes it look like there’s a correlation between the amount of non-coding DNA and organismal complexity. Fortunately, he counters it with a much more useful chart (that I’m definitely stealing for the next time I teach genetics) with no such bias.
And then Larry Moran tops Gregory with an even worse figure. I don’t quite understand it; maybe this distortion of the evidence to support progress, increasing molecular complexity, and the superiority of humans has roots in misunderstandings before my time, because my genetics and cell biology instructors in the 1970s sure didn’t promote this nonsense then. We were told even in those ancient days that the C-value paradox wasn’t a problem if you didn’t try to shoehorn mammals into a position at the pinnacle of evolution.
Maybe I just had really good professors. Thanks, Arthur Whiteley and Larry Sandler!
In a lecture yesterday the professor commented about an article by Levine and Tjian from 2003: http://www.nature.com/nature/journal/v424/n6945/abs/nature01763.html. He used terms like “more evolutionary advanced”, “more advanced on an evolutionary scale” and “higher eukaryote”. My respect for him got a lot less after silly stuff like that. I wanted to correct him, but I don’t have good enough argumentation to do it I fear.
you are right to point this out Tierhon, I’ve heard this sentence “more evolutionary advanced”, or equivalent, being used many times and each time I ask the speaker to explain, define, what he means.
I have still not been able to obtain a satisfactory answer.
Somehow, there is something completely wrong with the juxtaposition of these words.
Would appreciate it very much if PZ could give his take on this.
Oh, maybe I have really bad professors then. They reference of of my teachers at the Genome 2.0 article: Prof Mattick here at the University of Queensland.
I knew I had seen a very similar diagram (but expanded to cover more organisms) in his lectures. You can find it in this paper (John S. Mattick Nature Reviews Genetics 5, 316-323 (April 2004)DOI: 10.1038/nrg1321).
Or, it is also in this article: http://www.arxiv.org/abs/q-bio.GN/0401020
#2:”I’ve heard this sentence “more evolutionary advanced”
This was a common concept when I went to college back in the 60’s. It seems to me for the most part is dead now (as far as on-coding DNA, the Onion, which has more than us is a show stopper).
Later research and thinking about the nature of evolution basically showed that this idea, that we are headed somewhere that can be specified, to be false. So why is it still around? Well people who went to college at the time I did are still teaching, and while the good ones will have shucked it off, there may be some whose interest lies elsewhere and have not kept up. As a meme it is quite gratifying to humans because it automatically stuck us at the top fo the heap. The say “science progresses one funeral at a time” maybe we have not had enough funerals yet.
Why does this view as evolution as a climb towards progress have to define humans as the ultimate end? Maybe T. rexes are the pinnacle of evolution for all these people know.
A lot of the people who “believe in” evolution may only accept it because it fits in with their spiritual beliefs — life is an unfolding of Vital Essence manifesting itself through the progress of advancing Consciousness — or similar vague, fuzzy versions . And while this better educated, self-designated “enlightened” section of the general public might accept that the Great Chain of Being concept isn’t specifically in the theory of evolution, they think it can be easily grafted onto it, sort of like adding in “and God didit” at the end. But this is only because they don’t understand the actual mechanisms in evolution, and skim by on the surface narrative of “change over time.”
Getting rid of Teillard de Chardin-style vitalism may be just as shocking a move to spiritual views of the world as getting rid of special creation of species in a 6,000 year old earth was to fundamentalist views. But it will be a lot harder, I think, because this sort of argument against man as the end result of a progression is a lot more technical, and goes against more general intuitions.
Fits right in with the Nisbet-bashing thread, because the old view that humans are the crown and purpose of evolution is clearly of religious origin and influences the thinking even of biologists as distinguished as Miller, Collins and Conway Morris when that thinking is (as is inevitable) contaminated by their religious commitments.
Maybe T. rexes are the pinnacle of evolution for all these people know.
Giant Squid
Well said LaBonne.
I agree with you Steve, talking of us humans as the crown and purpose of evolution is a clear result of religious endoctrination.
Having said that, there is something very special about our lucky number, in the sense that it is very unclear to me if our future evolution will be governed by the same natural rules as the ones which have lead us to where we stand.
I decided not to use a full-on “text” for my cultural anthropology classes this semester: I hate them all, they’re expensive, and I spend the semester bitching about the thing on which students have just blown $80 or more. So I’m using something called “Core Concepts in Cultural Anthropology,” and the bitching has already commenced. In talking about the racist schema of unilineal cultural evolution from the laet 19th and early 20th century, the authors state: “These views are based on the fact that in biological evolution, organisms go from small to large and simple to complex over time.”
AGGGH! NO NO NO! Oh well, at least they only paid $30 for the privilege of experiencing my bitching.
I don’t see us an optimal species by any means, we have the childbirth problem yet, back problems due to bipedalism, a differential between males and females in the number and intensity of orgasms, etc. There is still a bit of work for evolution to do before we could be considered as finished. And of course we never will be. Evolution has no goals. The fact that we have a combination of a particular type of brain and the opposable thumb thing, we have on our own developed in to a narcissitic species who believe we can do more than nature can (and since we’re designers, we had to be designed, right?)
I read Larry’s article and Laelap’s article on how our growing understanding of evolution has led to the surprising notion that progression towards complexity is a red herring in understanding how it works. The shocking realization that “optimal species” are at the mercy of adapting to their environment is even more important than progression towards a goal, that is the thing that seems to shake the creationists the most. I don’t think I would feel at all “optimal” were I 2,000 feet down and facing an architeuthis. Yet, on my ground, I think I would be willing to take one on.
With garlic and butter.
I saw a talk by Mattick at Caltech about a year and a half ago, and I found it to be pretty compelling. The critiques of these stupid complexity graphs, while they certainly apply to dumbed-down science reporting and IDist quote-mining of his stuff don’t seem to be at odds with the parts of his work that seemed most interesting to me. On the other hand, I was feeling very much the “ignorant computer scientist” at the talk, so I took his word for it on most of the results, and a lot of more trained biologist types seem to take the “extraordinary claims call for extraordinary proof” view.
From my “ignorant computer scientist” perspective, it seems to me that because the ID weenies have decided they love the “junk DNA isn’t junk” stuff (and they’re helped by naive science journalists and pundits) a lot of pro biologists have a very strong resistance to the possibility that there are some important control systems for developmental and regulatory complexity embedded in the DNA that doesn’t code for actual proteins. Ironically, it seems like junk DNA was oversold as evidence for evolution, which is why the creationist wonks decided it was a worthy thing to attack. I saw no evidence that Mattick was some sort of creationist nut or even an overzealous crackpot… I’m curious about what Doddy (#3) thought of him as a teacher. The take-home lesson I got out of his talk was roughly “it’s been assumed that the DNA->RNA->protein model that works so well for E. coli is extensible to eukaryotes as well, but there are some reasons to believe that there may be other control systems in non-coding DNA that have not been studied specifically because we’ve been assuming without evidence that anything outside this model doesn’t exist.”
From a purely numerical standpoint, I’ve thought for quite a while that it would make sense that there is a more direct storage for blueprints in complex organisms– there around 10^11 neurons in the human brain, and only around 3×10^9 base pairs in human DNA, which seems like it would require something more information-efficient to contain the blueprint (although I can’t prove that the protein thing doesn’t have some emergent coding that happens to work, it seems improbable.) Since it is rather unlikely that it’s interleaved with stuff that codes for actual protein sequences, it seems entirely reasonable to me to assume it’s somewhere in the non-coding part (which may, of course, also contain some actual junk.) And it’s not at all clear to me that “the neo-Darwinian paradigm would be shattered by the discovery that all non-coding DNA has important uses that were placed there by THE ALMIGHTY LORD” is a logical consequence of investigating this hypothesis.
Where can I sign up for Genome 2.0? My current genome, Kallikak-Juke 3.6, is a real disappointment. I keep wanting to marry my cousin.
p.s. to Mike (#13): all the garlic and butter in the world won’t get rid of the ammonia taste. And anyway, you should have more respect for our cephalopod overlords. Vertebrate pride is hubris. Anyone know how many base-pairs in any cephalopod genome? Or whether there are sex chromosomes? Google says chromosome counts range from 28-56 in cephs, with octopoda and nautilus at the low end.
I have occasionally lapsed into the “more evolutionarily advanced” trap when attempting to discern between species that resemble earlier lifeforms versus ones of more modern advent–it is incorrect, and I always try to correct myself when I make those types of errors.
However, there does seem to be, especially when discussing classification systems and/or the “fossil” record a distinct sort of pattern throughout time. What would be the correct terminology to employ when first introducing high school students to concepts like prokaryote/eukaryote, or the use of the domain concept in classification? Remember, the goal is to get the ideas in place quickly, without getting bogged down in details.
And on a similar note: What is the consensus on evolutionary pressures on junk DNA? In other words, is there any survival advantage/disadvantage to having more or less of it? (I can see more junk as more raw material for variation–I can also see more junk as “noise” in the system as it attempts replication, growth, etc.)
Lastly, with what DOES the quantity of junk DNA correlate? Could it correlate with recency in the evolutionary process? (i.e. living things accumulate this stuff as generations go by?)
Gosh I wish I knew more about this stuff. . . especially since I’m teaching it!!
Oh, Mark, I respect cephalopods all right. The same way they respect me. As potential lunch. But I don’t think they consider humans all that much, whereas I am concerned for their continued existence. My point was that with all of the environmental variables here on Earth, there are no “pinnacles” of evolution.
Oh, and before anyone freaks out that I put fossil in quotations–I was just using the word as shorthand for all the types of preserved evidence we have, be they actual fossils (sedimentary castings formed in the cavities remaining from the decay of organic material) or other types of evidence of/from ancient organisms.
This is not true. “Pro biologists” know very well that there are control systems in non-coding DNA. Surprising new systems of regulation like RNA interference have been discovered recently. But these
1) still cover a very, very small percentage of non-coding DNA
2) in no way help the thesis that humans are special.
That certainly sounds inspiring, but what the heck does it mean, actually? Even if the regulation happens via a regulatory RNA, the information still flows according to the DNA->RNA->protein model.
I don’t get what you mean here. Do you think each cell is individually coded in DNA?
Isn’t there a difference between “progress” and “complexity”? Isn’t it conflating the two to automatically assume that if someone talks about the relatively high *complexity* of humans, that must mean they also think evolution is *progressing* towards higher complexity with humans at the pinnacle?
Isn’t it possible that humans *are* in fact more complex than (most? many?) other species, but this is *not* due to evolution *progressing* in this direction? The questions that need to be asked, it seems to me, are, first of all, “what IS complexity?” and then, “ARE humans more complex than other species” – i.e. does our subjective assessment of complexity by eye correlate with actual biological complexity – and then finally, “why”. I don’t think this is the same thing as asking whether evolution has been progressing towards a final goal of humans, which is obviously not the case.
Or am I wrong, have these questions already been answered? Has it already been scientifically determined that humans are no more complex than other species?
And some way to remove the stinking ammonium salts they are filled with for buoyancy…
These “parts” are interleaved. The genes are strewn around randomly, as far as we can tell. (Stupid design.)
Being prokaryotic is the normal state of affairs, being eukaryotic is an innovation.
The domain is just another rank, like class, phylum, and kingdom. There is no concept.
Total genome size correlates to cell size. Volume-to-surface ratio of cells is sometimes important — sometimes.
Impossible — all known life has a common ancestor.
This (a natural cast) is just one form of fossil. It is more common, especially with bones and teeth, that the actual material of the hard parts is preserved and the micro-porosity filled in.
One proposed measure of complexity after another has failed…
But you will be pleased to read that you’re in good company with your idea. It’s an important part of Gould’s book Full House.
And some way to remove the stinking ammonium salts they are filled with for buoyancy…
These “parts” are interleaved. The genes are strewn around randomly, as far as we can tell. (Stupid design.)
Being prokaryotic is the normal state of affairs, being eukaryotic is an innovation.
The domain is just another rank, like class, phylum, and kingdom. There is no concept.
Total genome size correlates to cell size. Volume-to-surface ratio of cells is sometimes important — sometimes.
Impossible — all known life has a common ancestor.
This (a natural cast) is just one form of fossil. It is more common, especially with bones and teeth, that the actual material of the hard parts is preserved and the micro-porosity filled in.
One proposed measure of complexity after another has failed…
But you will be pleased to read that you’re in good company with your idea. It’s an important part of Gould’s book Full House.
It’s an important part of Gould’s book Full House.
In which he argues for the “modal bacter,” pointing out that, to a first approximation, all life is and always has been prokaryotic. He conceptualizes complexity on a graph, with bacteria and archaea reprsenting “the left wall,” and eukaryotic complexity as a “tail” to the right. A frothy and insubstantial scum on the prokaryotic pond, you might say, barely distinguishable from statistical noise.
Given that there are roughly 1018 insects and 1030 bacteria on the planet, it is clear that by at least one measure, complexity doesn’t correlate with success or best strategy.
We see that an increase in complexity can be a positive trait in many ecological niches, else there would be far fewer or no complex organisms at all. Only when there is a net cost to becoming more complex will a ceiling become apparent.
It seems to me that there is a ratchet effect that favours more complexity rather than less when a mutation is selected for, a creeping featurism effect that responds to challenges by adding a new ability on top of existing ones.
And indeed Gould says it is statistical noise.
A cost always exists. Whether it’s a net cost depends, obviously, but…
…in that case we’d expect to see a driven trend. We don’t.
If anything, we see a recurrent driven trend towards simplification (parasites), but I don’t think anyone has looked at how driven even that is.
And indeed Gould says it is statistical noise.
A cost always exists. Whether it’s a net cost depends, obviously, but…
…in that case we’d expect to see a driven trend. We don’t.
If anything, we see a recurrent driven trend towards simplification (parasites), but I don’t think anyone has looked at how driven even that is.
Thanks, Marjanović. I appreciate the refinement on my (weak) off-the-cuff definition of fossil, as well as the cell-size-to-junk-DNA correlation (something I had not thought of).
I guess what I meant by “domain concept” is that, from what I can tell, the 5-kingdom system I studied as a student has now another layer of classification seemly grafted atop it–and I suppose the intent behind this is to further indicate our best guesses about the “branching” of the “tree of life.” Or, to work my way from the tree analogy (suggesting there is a top somewhere), perhaps I should say the domains seem to be there to emphasize the relative degree to which evolution has wrought both physical and genomic change in each group since the first organisms came into being. This is the idea behind having domains==domain concept.
Or maybe I’m making all this too complicated: There is an apparent effort to formulate the classification system to reflect evolutionary history–a sort of “family tree”. How does one explain this effort (and more importantly, bring some sense to the way in which we classify) without conjuring notions of “more advanced” or “of higher complexity” (again, without getting bogged down in details or terminology)?
It sounds like my representation of Mattick’s work didn’t really get the idea across (I *did* say I was an ignorant computer scientist, though…) By interleaved I meant “contained in the triplets that encode the actual amino acid sequences somehow”– I do know that these genes are distributed more-or-less randomly across chromosomes.
regarding #20 (windy):
Mattick’s hypothesis, as I understand it, is that it’s at least worth investigating whether the item 1 is more of an artifact of not knowing how to interpret the non-coding part, and really what’s been found is the tip of the iceberg. I didn’t make any claims about humans being special, so I’m not sure if that’s addressed at me or other comments. Mattick pretty much thinks that non-coding stuff carries a lot of information in eukaryotes in general.
um, I guess it means that I didn’t give a very clear explanation. Essentially, the sort of thing he wants to look at is that there is a whole lot of DNA that’s translated to RNAs that aren’t ever connected to ribosomes, and that appear to lack essential elements to be directly transcribed to proteins. These have traditionally been ignored or written off as “junk” to some extent, or at least “we don’t know how this fits into the system, so we won’t talk about it much.” Mattick used a metaphor that the DNA->RNA->proteins system is analogous to the basic components of a building being assembled (like 2x4s or drywall or air conditioners or electrical outlets: basic generic building blocks) and posits that there is also a separate system that specifies something more akin to the architectural blueprints for the building: this is the layout of the organism, and how things are connected. Since a whole lot of base pairs in the genome are transcribed to RNA and never translated to proteins, it seems worth asking the question of whether this RNA is really just junk that’s transcribed and thrown away, or whether a lot of it may be important in some way we don’t understand. For example, it might be DNA->RNA->some RNA-based control system->regulated transcription of other genes->DNA->RNA->protein. If that’s “the same as DNA->RNA->protein” in some sense, then maybe we’re saying the same thing from different points of view, but I can imagine all sorts of things that this RNA might do, like modify the mRNA or tRNA that’s going to be turned into protein, or activate other genes, or be a ribozyme that reconfigures proteins or some other enzyme task, or interacts with ribosomal RNA… going back to the metaphor, he pointed out that the kinds of changes you’d see in a mutation of the blueprint would look different than the ones in the components– if the chair factory were broken, all the chairs would look wrong, but if the blueprint were off, there might be 7 chairs around a conference table instead of 8, but if you were just looking at “are the chairs normal or mutant” they would look fine.
Mattick also put forth the hypothesis/metaphor that the protein production part was similar to analog systems, and that this hypothetical blueprint stuff was more akin to digital systems. I’m normally actually pretty skeptical about applying computer terminology to biological systems, but in this case I found it interesting, at least.
Not at all. Just that the whole body has to somehow be described in 3×10^9 base-pairs, and yet the brain is a very complicated structure composed of 10^11 neurons, and the rest of the body is pretty complicated also, of course (I didn’t bother to look up how many cells there are total, I just happen to know the brain)… what I’m getting at is that if someone were to try to write a book on how to make a brain, it would take a whole lot of information to describe how the brain is wired together, yet developmentally, somehow, the brain (and the rest of the body) gets wired together in such a way that it very reliably is able to be connected in these complex ways and is generally robust and functional. If the system is primarily based on protein transcription, it’s really not very information-dense in terms of specifying the layout of the brain, since most of the information is going into things that specify all the nifty geometry that gets active sites of enzymes and membrane proteins and such to the right place, and there’s a whole lot of stuff that’s just scaffolding to do that. Although there are often pretty complex structures that can arise from simple rules, I’m skeptical that the varieties of living creatures that we see are constrained to only have structural complexity created by iteratively applying simple rules like this– it seems far more likely that there is some sort of information storage at the sort of blueprint level that Mattick proposed. He also discussed some work by a mathematician that purported to show that prokaryotes were right at the limit of what a mathematical model predicted for where the complexity of “analog” regulation would break down in some way (I’ve forgotten the details, and there wasn’t enough information in the talk to assess whether the model might be flawed… I think it was a combinatoric argument of some sort… I’ll see if I can find my notes.)
I reiterate, though, that I’m not qualified to advocate this position particularly at all, I’m just perplexed because it seems like an entirely reasonable line on inquiry, but there are a whole lot of biologists who seem to be vehement in arguing against investigating whether there may be more to the noncoding DNA than we understand. It’s certainly understandable that it’s really, really annoying that creationist kooks and bad science writers are jumping on this to write stories that suggest that biologists are fundamentally wrong, and I apologize if I’m coming across as in the same camp. My impression is that Mattick and a few other credible-seeming people are saying “the established models of genes and regulatory systems are right, and necessary, and really good work, but there may be a very big picture lurking in the shadows that it could be very worthwhile to investigate as well.” Maybe he’s a convincing crackpot and I’m a gullible ignorant computer scientist, but I’m pretty capable of understanding biological arguments, and I haven’t seen many solid ones that say that Mattick’s take is fundamentally flawed in the same way as the creationists and bad science writers is, so I’m either looking for an understanding of why I should think Mattick is proven wrong, or I’m inclined to think that ire at pinheads is potentially causing a “throw the baby out with the bathwater” issue.
I’d certainly love to hear how some of you more-qualified-than-I-am folks dismiss Mattick’s work and why, though, because although I dislike a lot of the mainstream media presentations, my sense is that there’s at least a possibility that some of the non-coding DNA is pretty important in some sort of an information storage/ regulatory system sort of way… Mattick seemed to be willing to look at the hypothesis skeptically and accept evidence that would exclude it, and the more he looked the better it looked, but maybe he’s just a good salesman for a dumb idea. I’m certainly interested in being convinced of that possibility, but the “dumb creationists and ignorant journalists present a flawed version of it” argument just seems to (effectively) prove that creationists and a lot of journalists are dumb. I reiterate also that I don’t intend to say that humans are special, or even that metazoans are special, particularly, or that counting base-pairs or genes or introns or whatever is a very good measure of “complexity.” And I agree with the point a lot of people seem to be making that ranking things on some scale of organismic sophistication is silly. There may be some sorts of metrics for complexity or information theory that might be useful here, but I’m not sure what they are (the Avida guys like to use thermodynamic and entropy arguments a lot, which I’m not so keen on both because I’m never sure what their underlying assumptions are, and because statistical mechanics makes my head hurt.)
Part of the problem, at least from a physics viewpoint, is that no single measure can capture all characteristics.
I just learned from Moran that Taft recently proposed an “organism complexity” “broadly defined as the number and different types of cells, and the degree of cellular organization.” That sounds like a neat first order approximation, but there are many traits and behaviors left to capture.
And after you have a description, you have to find out if it is meaningful (for prediction)…
Part of the problem, at least from a physics viewpoint, is that no single measure can capture all characteristics.
I just learned from Moran that Taft recently proposed an “organism complexity” “broadly defined as the number and different types of cells, and the degree of cellular organization.” That sounds like a neat first order approximation, but there are many traits and behaviors left to capture.
And after you have a description, you have to find out if it is meaningful (for prediction)…
Mark Montague:
IANAB, but the TR Gregory, specializing in genome size evolution claims:
Mark Montague:
IANAB, but the TR Gregory, specializing in genome size evolution claims:
Mark Montague:
Simple rules like this?
There are many more posts on Pharyngula explaining how complex and jury-rigged evolutionary regulatory systems are.
Further, if you don’t succumb to evo-devo, PZ will probably come and eviscerate you with his beak!
Mark Montague:
Simple rules like this?
There are many more posts on Pharyngula explaining how complex and jury-rigged evolutionary regulatory systems are.
Further, if you don’t succumb to evo-devo, PZ will probably come and eviscerate you with his beak!
Oh, I forgot – the last link is to Drosophila. The texts notes that the fly, probably because of its rapid development where much happens at once, has traded off general development modules for hardcoded ones. This fly has a more complex development than most regarded as gene interactions.
Oh, I forgot – the last link is to Drosophila. The texts notes that the fly, probably because of its rapid development where much happens at once, has traded off general development modules for hardcoded ones. This fly has a more complex development than most regarded as gene interactions.
Again, completely wrong. People are investigating it actively, not just Mattick. If Mattick doesn’t want people vehemently arguing against his ideas, he should note that his silly figures where he always puts humans on top are bound to be vehemently argued against because they are wrong.
The strength of the human brain is that a lot of it isn’t hard-coded and it learns from the environment as it develops. But even if a lot of it is hard-coded, think of the difference between a human brain and a chimp brain. Somehow, humans manage a lot more neurons with the same amount (or slightly less) non-coding DNA.
For starters, check out TR Gregory’s “onion test”.
Nobody is against looking for more functionality in non-coding DNA. But I think it’s better not to tie this search to very fuzzy assumptions about “complexity”.
negentropyeater: Assuming we don’t kill ourselves off by ecocide, we may well become far more dependent on our technological environment. Aside from that, I suspect our future evolution will be mostly directed toward intraspecies competition. The results might be damned peculiar…. ;-)
rubberband@19 and following commenters:
Complexity is a substrate for new capabilities, and a complex organism can find and defend niches which a simpler form can’t.
Parasites are something of a special case, but the basic reason some of them can shed complexity is that once infected, there’s often not a heck of a lot the host can do about the infection. (The classic example would be tapeworms.) So, parasites of those sorts can afford to focus on feeding and replication.
My comments all end up in moderation but hopefully they will turn up later. I see Torbjörn already provided some links to TR Gregory, that’s a good place to start.
I would appreciate it if you can explain how protein regulation is significantly more “analog” than RNA regulation, I don’t immediately see it from the bits you quoted so far. Note that bacteria self-organise in biofilms etc. so one should be careful of making pronouncements of a bacterium having reached maximum complexity.
Mark,
You forget about how much information is embedded in the environment. Organisms aren’t created ex-nihilo, but are absolutely dependent on a fairly narrow range of environments to properly mature.
Try to build a computer with just the technical specs and a stick – no other people, no technology.
On the other hand, there was a ENCODE Nature 477:799 (2007), where they seem to be suggesting that there is a pattern storage engine in the junk DNA.
Yes, but “it would make sense” is not a compelling argument. There are a lot of things in the universe that don’t “make sense” which happen to be true. Things which are very large or small, or happen over a very short or very long period of time, consistently run exactly opposite to the intuition of an ape that stands about five odd feet tall and lives about seventy odd years.
As to the brain, every one we’ve so far examined is different, so it seems there’s at least ninety billion or so ways to make human brain, with new permutations showing up with increasing frequency. This does not seem to indicate the necessity of any but the most basic blueprint.
Just to clarify, I didn’t intend to suggest that Mattick is the only person working in this area, just that he’s outspoken, was an early proponent, and I was impressed by his talk. I read some of the ENCODE results published a few months ago, and found that it seemed relatively consistent with the sorts of things that Mattick had said, and I also mentioned this stuff being presented as new to a genetics guy, who said that this stuff was expected since the 70s or 80s by a number of folks, including Eric Davidson, so it’s not all that groundbreaking. I’m just finding a lot of the “arguments against” the noncoding DNA seem to address a pretty naive version of it.
I’m kinda feeling over my head in that I really don’t remember the details of Mattick’s stuff that well, and I understood the ENCODE stuff even less, so I was hoping someone who knows the details more than I do would step in and explain “Mattick says exactly X, Y, and Z, and the arguments for and against are A, B, C.”
Some vague attempts to explain why I’ve said what I’ve said so far, with the caveat that some of them will be naive:
* it’s not obvious that ncRNA stuff is inherently digital, Mattick just thinks that a more digital system that has “on/off” rather than “upregulate/downregulate but the levels can vary continuously” is a plausible thing to look for, and it might be found in the ncRNA that seems to have some things in common with what one might expect in such a digital system.
* there are all sorts of places information could be hiding in the genome and the environment, but I’ve often been perplexed at the amount of repeatable organization that emerges in metazoan development, and it seems implausible to me that the sorts of regulatory networks that have been brought up as examples of patterning are the whole story (although they’re obviously an important part of it, too), and I don’t see much evidence that the environmental information is the main driver, either (it seems to me that a lot of biological systems don’t show the fragility that’s often associated with using environmental interactions and emergent complexity from simple rules frequently lead to)… ncDNA (or whatever you’re supposed to call the DNA that is transcribed to ncRNA) is something that is actually copied with high fidelity and transcribed and left to float around for a while in the cytoplasm, which at least seems like a very plausible place to look for extra information storage and some mechanism to access it… I certainly believe that there’s plenty of room for random junk in the genome as well, but I have found a lot of the arguments for the lack of function sound a lot like “absence of evidence is evidence of absence” sorts of views…
here’s a summary letter that doesn’t need a subscription to access the PDF and has some references:
http://www.psi.utoronto.ca/~frey/tcb/papers/RNAFunction.pdf
This one is a lot more detailed, and covers some of what he mentioned in this talk:
http://hmg.oxfordjournals.org/cgi/reprint/15/suppl_1/R17.pdf
and this one is specifically about the argument of scaling and limits of prokaryotes:
http://arxiv.org/pdf/q-bio/0312021
Again, though, I’m mostly just doing a search on the Mattick worldview, while it certainly seems like that’s disputed to some extent, and I’d be interested in understanding the “Mattick is wrong” viewpoint better as well… I have to admit to a certain fondness for the notion that there’s more to things like introns and other non-coding genes than meets the eye, but my fondness isn’t, of course, particularly related to reality (and it’s certainly not correlated with a belief that it wouldn’t have arisen through evolution: I find it plausible that some eukaryote stumbled on a system that allowed better concise encoding of some regulatory information, and it conferred a large fitness advantage.) I also find some of the arguments for the “junk” DNA being useless sound a bit circular: the parts of the genome that aren’t exons seem to not follow the evolutionary and statistical trends we understand for exons… that doesn’t prove they’re disfunctional, it just proves that using the exon model to understand them isn’t a good way to see if or how they’re functional. I’m also sort of perplexed that there don’t seem to be any experiments that attempt to cut out large amounts of the supposedly useless DNA and either show that the organism can develop fine without it, or not. I assume this is because it’s a lot harder to do this in nuclear DNA than in plasmids or something, but it seems like it’d settle the argument pretty quickly: “we chopped out all this stuff that Mattick hypothesizes is important, and it didn’t hurt this euglena/flatworm/fly at all.” I sort of expected this sort of result from this post: http://scienceblogs.com/pharyngula/2007/06/its_junk_get_over_it.php but really didn’t find much to convince me that the classification of junk DNA as non-functional is based on solid evidence… I’m not saying that the evidence isn’t there, just that I didn’t see it, and I do see a lot of discussion degrading into a competition where junk DNA has a special place in the hearts of those who hate creationists, because it’s useful for smashing creationists arguments, and creationists, in response, love to use anything that was thought to be junk that isn’t as an exaggerated defense against this (fueled by science reporters who love to report “accepted scientific views have been turned upside down by this new discovery”) and this doesn’t appear to leave a lot of room for legitimately asking whether some of the believed-to-be-junk DNA actually could have some function, and whether the metrics for how it’s decided “this is junk” are necessarily valid. A lot of the folks who attended this: http://www.cbcd.caltech.edu/replication/index.html (where I was a lurker) have actually made systems that have made digital-ish systems out of nucleic acids (or other things) that show interesting information storage behaviors… I’d think that if some early organism stumbled on a similar system, it’s likely to have provided it with a new tool that increased its fitness and enabled it to explore new ecological niches.
Oh, you were looking for a definition of fossil? Try “remains of, by, or from living organisms”. That includes footprints, burrows, nests and so on.
This is all just within traditional classification: it’s just the addition of another rank on top to slightly reduce the fact that there are never enough ranks.
Oh yes.
I don’t see why any such notion is necessary — or even easy to get in. Talk about diversity. And read Full House :-)
(I like to say that Full House explains evolution by baseball. After reading it, I still haven’t understood baseball, but I’ve understood evolution.)
Any organism can find and defend at least one niche that any other form can’t.
But an extremely common one. Not to mention the epiparasites (parasites of parasites… of parasites of parasites…).
Parasites, tapeworms included, jump through lots of hoops to escape the immune system.
Some say that the reason why allergies are so much on the rise is that our immune system is by default hyperactive — because we have adapted to the constant presence of all those parasites that suppress our immune system.
Oh, you were looking for a definition of fossil? Try “remains of, by, or from living organisms”. That includes footprints, burrows, nests and so on.
This is all just within traditional classification: it’s just the addition of another rank on top to slightly reduce the fact that there are never enough ranks.
Oh yes.
I don’t see why any such notion is necessary — or even easy to get in. Talk about diversity. And read Full House :-)
(I like to say that Full House explains evolution by baseball. After reading it, I still haven’t understood baseball, but I’ve understood evolution.)
Any organism can find and defend at least one niche that any other form can’t.
But an extremely common one. Not to mention the epiparasites (parasites of parasites… of parasites of parasites…).
Parasites, tapeworms included, jump through lots of hoops to escape the immune system.
Some say that the reason why allergies are so much on the rise is that our immune system is by default hyperactive — because we have adapted to the constant presence of all those parasites that suppress our immune system.
First of all, did you look at TR Gregory’s blog through the links above?
Yes there are, look up “Megabase deletions of gene deserts result in viable mice” in Nature.
More recently, even a few ultraconserved non-coding DNA sequences (not junk) were removed and the altered mice did well anyway! Nobody can explain this result very well at the moment, so I expect we’ll hear some interesting things in the next few years.
Why? A hardcore adaptationist would say that he never expected that much useless DNA to be retained without a function. Evolution does not require outright junk. And the creationists still have to explain why there are retroviruses and pseudogenes in identical locations in the genomes of related species.
I think a problem is that when people find a few transposons, or retroviruses or microsatellites or whatever, doing something useful, then they sometimes go and say “maybe the transposons/retroviruses/microsatellites are there for a reason”. This is completely unwarranted. We would expect an “unwanted” source of mutation to now and then produce something useful that is then retained.
What were these sequences?
What were these sequences?
What were these sequences?
here’s the article on PLoS Biology: Deletion of ultraconserved elements yields viable mice. But they only looked at four so far.
Thanks, Windy (and others)
The megabase deletions mouse is pretty much what I was curious about. I read the TR Gregory blog and found it interesting, but it seemed to involve a lot of “this is how it is, trust me” assertions rather than the supporting details, which is probably part of some of my frustration with this issue… the best argument I know of against intelligent design is that in biology, the way you think it would make sense for it to work is never how it turns out to work, so reverse-engineering a biological system has to go through many more iterations of “I had flawed assumptions” than reverse engineering, say, a computer file format. This “hard to reverse engineer” seems to apply to simple evolved artificial-life computer creatures, too. However, a lot of the arguments for what is and isn’t functional in the genome seem to me to be lacking the feedback on whether the assumptions are valid… I can think of hand-waving arguments for how ncDNA that has some bizarre function might be strongly conserved or not conserved at all or anything in between by various metrics… if one applied the same techniques used to argue this stuff to snapshots of a computer’s RAM, for example, depending on whether you were looking at a CDDA audio file, an mp3 audio file, a compressed image, and uncompressed image, an executable program, or a the text of a document, or a spreadsheet, you’d see all sorts of different statistics about the sequences of bytes. I normally hate applying computer analogues to biological systems, so please take this with every caveat you can imagine (e.g. no supernatural computer programmers producing DNA sequences, no belief that cells are Turing machines, etc.) To take a ridiculous example, suppose that some cell evolved the “zip” algorithm to more efficiently store some information. If some biologist were looking at the zipped version of the genome of two cells with a single point mutation, the zipped versions would differ at some huge number of loci, and would seem to be completely uncorrelated if one didn’t know the underlying zip algorithm. And the “this is the start of a zip file” tag would be ultra-conserved. Obviously, this has nothing to do with real biology, but it’s intended to illustrate that a lot of the arguments for what’s junk or not seem to depend on what’s conserved, or what drifts, or that sort of thing, and that causes a chicken-and-egg problem: if some of this stuff is doing something no one expects (in biology, that’s pretty much the norm, right?) then one should be very careful about what assumptions and tools one applies to decide how to interpret it.
Here’s a simplified model of what I’m talking about, the wikipedia page on introns seems to be a pretty accurate representation of what’s taught in overviews of biology: introns get spliced out of the mRNA before it’s translated. It doesn’t mention at all what the fate of the spliced-out parts are… it seems to be quietly assumed that some enzyme or something just breaks it down immediately, or even “it’s not translated by the ribosome, so who cares?” Searching google for “intron splice fate” does produce some interesting results, but, e.g. an abstract that says “Given the importance of intron turnover, it is surprising that there have been no reports on the half-life of introns from higher eukaryotic cells.” (leaving out URL so I don’t need to wait for moderator.) That 1999 RNA abstract was emphasizing that these snipped-out introns tie up valuable nucleic acids, but it suggests that people hadn’t even studied how long these RNA strands hang around, let alone what they might be doing during that time… maybe I’m naive, but that sounds like an “if it’s not made into a protein, it can’t be doing anything” bias that writes off everything as a spandrel by fiat rather than evidence. Admittedly, I’m doing a typical quotemining thing by citing that, since I’ve only read the abstract and glossed over the parts of that that I didn’t understand, but I did it because it really ties into what’s been bugging me: Mattick seems to be saying “maybe we’re wrong in our assumptions about what’s important in the non-coding areas, and we’ve been ignoring areas as non-functional based on questionable assumptions of how to tell if things are non-functional.” I may not be qualified to intelligently critique the criteria for concluding that some span of DNA is non-functional, but Mattick and various others appear (not just to ignorant me or ignorant creationists, but to reviewers at Science and Nature) to be qualified to bring up these issues. And again, I don’t particularly mean to single out Mattick as some sort of hero or maverick genius, I just went to this talk he gave, and found that he seemed to make a lot of sense, at least at the level of “it seems like this is a hypothesis that should be studied experimentally rather than dismissed as obviously wrong.” So I’m curious about why other people I respect, e.g. PZ, write blog entries like “It’s Junk. Get Over It.” and seem to be convinced that asking the questions is beating a dead horse.
Oh, come on. He raises valid questions (why does an onion need so much DNA, is it five times as complex as human?) and points out that Mattick’s selection of organisms for his “complexity scale” is questionable.
Here’s another flaw: Mattick has repeatedly said that it’s wrong to assume that gene regulation and genomes in prokaryotes and eukaryotes function the same way. But when he draws a connection between noncoding/total DNA ratios and organism complexity, he includes prokaryotes with eukaryotes – using only eukaryotes, the correlation would disappear!
Lots of people are looking for functionality in parts of noncoding DNA, nothing wrong with that. But I think you are ignoring the sheer volume of gunk out there. The most out-there estimates could be that a fifth of human DNA is functional -that still leaves 80%!
Here’s some more from Genomicron posted today:
If you want to know more about what these “mechanisms of accumulation” are, look up mutation processes in microsatellites, pseudogenes and transposons.
This nutbar… One of my favorite things was making fun of various Americans attempting to remove evolution from school… now I can’t do that with the same sense of fun, as this idiot is setting up in my back yard. *sigh*
Thanks again, Windy. Yeah, I think the newer Genomicron post is a lot better for what I’m curious about. I don’t find the onion argument terribly compelling– it’s something of a red herring, particularly from my perspective, which is that some ncDNA probably is junk, and some probably isn’t junk, so if the onion either has disproportionately more actual junk, or doesn’t use its non-junk very efficiently, it doesn’t really bother me. The real thing I’m after is that there seems to be an attitude of “Assume non-coding genes are junk until proven otherwise” rather than “assume we just don’t know, and then look at the non-coding genes to see.” Sure, lots of things in the genome clearly arose from understood processes (didn’t someone recently reconstruct an extinct retrovirus from inside some genome?)
But there seem to be some plausible-sounding arguments in the other direction, too: why do we bother to transcribe all these junk genes to RNA? Shouldn’t there be selection pressure to stop doing that, since it presumably helps retroviruses and has a metabolic and resource cost?
Something I haven’t heard is where all these estimates for “what percentage of the genome is functional” come from… that always strikes me as similarly implausible to “we only use 10% of our brains” (and argument often repeated by people who only use 10% of their brains, but rarely by neurobiologists.) Identifying actual exons appears to be rather well-understood. But the “estimating non-function” part rarely seems to be discussed at a non-academic level, except to assert “scientists say no more than 20% of the DNA is functional.” I’m interested in what the assumptions are for deciding what’s “junk” ncDNA as opposed to “we have no idea if this ncDNA is functional or not.” I’m particularly suspicious of claims like “there is indirect evidence that most of it is not [functional]” and “…these [accumulation mechanisms of junk] could lead to increases in DNA content independent of benefits for the organism, or even despite small detrimental impacts, which is why non-function is a reasonable null hypothesis.” (from the link in #29 above.) There may be great evidence for those claims, but I haven’t found any where I can see what the data and assumptions are that lead to those conclusions. Of course, the mice with mass ncDNA deletions are, in fact, a great example that I hadn’t been aware of, although it still seems somewhat preliminary (deleting 10^6 base pairs out of 3×10^9 would be 1/3000, as I cheated by using the human # which doesn’t really prove that 95% of the genome is junk until you can prove that it’s got a whole lot in common with the 1/3000 that was chosen.)
Anyway, thanks, and I should probably stop cluttering up PZ’s comments with my questions…