|
Author
|
Topic: The Evolvability of Redundancy
|
John Bracht
Member
Member # 5
|
posted 01. August 2002 00:10
I've been talking with a friend of mine, Ryan Huxley, who is a structural engineer and who has written an article on redundancy and design, from an engineer's perspective. This is a topic that I've thought of occasionally with regard to ID and evolution, and I wanted to post some of our ideas and intuitions in hopes of getting some feedback from both biologists and engineers.
Basically, Ryan and I agree that redundancy would be very difficult for Darwinian evolution to produce. The reason is that a redundant system provides no immediate selective advantage. After all, the very definition of a redundant system is that it exists in addition to another system which already provides the same functionality. The redundant system exists only for the prevention of failure in a key system, but it provides no immediate, positive, selectable advantage. In other words, even if an organism somehow evolved a redundant system, there would be no selection pressure to preserve it and it would likely be lost.
In fact, the Darwinian mechanism should favor the loss of redundant systems because the cost of building a whole, separate system that provides no immediate advantage would tend to penalize those organisms that carry them. Therefore, we should expect an evolutionary tendency away from redundant systems.
Furthermore, there are some interesting (and important) subtleties. For instance, the redundant system incurs a maintainence cost just like the primary system. If a redundant system is not maintained, it will not be available to fulfill its function if/when the primary system fails--and it will be no better than having no redundancy. So there's a manufacturing and maintainence cost to redundancy that makes it quite expensive. That's probably one reason we only see it in systems that absolutely, positively, must not fail (like life-support systems on the space shuttle, etc). Again, we're seeing a strong evolutionary pressure to jettison redundancy even if it does arise by some non-telic process.
In attempting to attach some probabilities to the probabilities involved with redundancy, I recalled some ideas from population genetics. It turns out that there is a theory of evolutionary change called the "neutral theory" which holds that most evolutionary change is actually driven by neutral changes, not by positive selection. Anyway, one of the outworkings of this theory is that even mutations that provide selective advantage are often (and usually) lost from a population. The way the math works out, if a mutation provides a fitness advantage of S (percent) for the organism, relative to its peers, there will be only a 2S probability that it will be maintained in the population--and a 1-2S probability that it will be lost. Thus, if a mutation arises that confers a 2% fitness advantage to an organism (relative to the other organisms in the population), then there is only a 4% chance that the mutation will be retained in the population and a 96% chance that it will be lost, just by chance genetic drift.
It seems that this basic framework applies nicely to redundancy. To understand this, think again about the two systems in a redundant system. I'll call them the "primary" system and the "redundant" system. As long as the primary system doesn't fail, there is no need for the redundant system. So there will be absolutely no selective advantage for having the redundant system. I like to say that the primary system "shields" the redundant system from the view of natural selection. However, imagine a primary system that is not so reliable, and it fails 2% of the time. This means that 2% of the time, the organism with the redundant system will survive whereas its peers will die (because they lack the redundant system). The implication is that this particular redundant system confers a 2% fitness advantage over non-redundant systems. A redundant system coupled with a primary system that has a failure rate of 2% has a 4% probability of being preserved in a population, and a 96% probability of being lost. A primary system's failure rate corresponds to the fitness advantage provided by a redundant system (when coupled with the primary system in question).
It gets even more fun. Imagine that our redundant system is a multi-part, irreducibly complex system. Imagine, furthermore, that it operates in a unique way from the primary system and cannot just be generated by copying the primary system's DNA code. Now, in order to calculate the probability of that redundant system's arising, we must not only factor in the probabilities of particular duplication/mutation events (which must happen in one fell swoop to get functionality), but we have a "shielding factor", due to the primary system, that reduces the probabilities even further. So imagine that we calculate a probability of a given system's occurance by chance alone, and that system is redundant to a primary system that only fails 1% of the time. That gives us a shielding factor of 98%, so suddenly the probabilities are 98% less than they would otherwise be. And if there are more than one redundant system, there will be additional shielding factors piled on top of each other. One can see the probabilities headed sharply downward.
This leads to some obvious and interesting questions: how does this all apply to biology? Can we find redundant systems in biology? Is redundancy very common? How much redundancy (in other words, what level of redundancy, i.e., how many parallel redundant systems for a given overall function)? Is the redundancy given by genes homologous to the primary system (in other words, is it possibly due to gene duplication), or is it due to unique and different genes?
I want to address the inevitable gene-duplication argument. Certainly, merely duplicating existing genes is a good way to get redundancy. My analysis above assumes two independent systems. But even a redundant system that arises by gene duplication (which, as far as we can tell, is a relatively common event) has the same negative selective forces described above: the manufacture and maintainence costs that should favor disabling or formation of pseudogenes from the duplicated genes.
In fact, some organisms have many duplicated genes. The frog, Xeopus laevis, appears to have tetraploid genes (its entire genome seems to have been duplicated twice). However, the amazing thing (and the mystery to biologists) is that relatively few of those duplicated genes have been disabled by mutation--many of them have been preserved somehow in a functional state (as far as we can tell). So here is a good example where much of a genome is actually redundant, and is maintained in a functional state. From a Darwinian standpoint, this is a mystery. What possible selective advantage could these genes be providing? (I should add that these genes have retained great sequence similarity to their "parent" genes and thus likely are still providing the same, hence redundant, function). I was talking with a co-worker in the lab today who gave me a paper on the Arabidopsis thaliana (the most well-studied plant in biology) that reports finding that over 50% of the genome is duplicated and, apparently, redundant. She pointed out to me that the real mystery, from an evolutionary standpoint, is why those genes are still there and are still functional. Why haven't more (or most) of them become pseudogenes? There isn't an easy answer.
Another interesting point: redundant systems are very hard to study in biology. In fact, this same co-worker in the lab told me she had spent the first 3 years of her doctoral work trying to isolate a particular mutant, and she was unsuccessful. Why? Because it was redundant. That means that when she tried to generate mutants, there was no observable effect on the plant and hence nothing to select for (because the redundant system was compensating for the mutations she was inducing in the primary system). She pointed out that the redundant gene could have been almost anywhere in the genome and could well have had a totally unrelated sequence to the gene she was trying to knock out. She was unable to identify the redundant gene(s). So apparently homology is not necessarily enough to identify genetic redundancy--the redundancy can exist via (apparently) unrelated genes, in an independent manner as I've outlined earlier.
And just today I was reading a paper which says
quote:
However, the genetic analysis of mammalian MAP kinase pathways, including the p38 MAP kinase pathway, has been complicated by embryonic lethality and presumed redundancy [reference deleted].
(emphasis added)
(source: Kim et al. Science 26 July 2002;297:623-6)
The fact is, if you have a redundant system, it is going to be very robust to disruption. However, much of modern cell biology and genetics relies upon disrupting biological function, then studying precisely what was disrupted. This implies that there may be many redundant systems out there that have not been identified simply because they are redundant and have never been disrupted. If redundancy is a common feature, then we would expect future research to reveal many novel biological pathways that have been overlooked till now, as biology has (to this point) focused on the few relatively easy-to-disrupt non-redundant systems. The group that begins looking for ways to detect redundancy would have a big advantage in future research. This may be a promising avenue of research which is specifically prompted by ID ideas (noticing how intelligent humans use redundancy for important systems, can carrying that insight over to biological systems). Furthermore, a Darwinian mechanism, it seems, cannot adequately account for either the origin or maintainence of redundant pathways and thus this avenue of research would be much less likely to be explored from a Darwinian perspective.
So, I hope this post is thought-provoking, and in particular I would like to hear from biologists with ideas about the following questions:
1. What are some concrete examples of redundant systems in biology?
2. What types of redundancy exist in biological systems? Is it "unique" redundancy (the redundant system is totally different from the primary system) or "cookie cutter" redundancy (the type of redundancy that results from copying the primary system)?
3. Can we get a handle on the probability of failure of the primary system in cases of redundancy? What do these probabilities tell us? Are any redundant systems irreducibly complex?
4. Are my ideas about a "shielding factor" and population genetics appropriate? Does this seem like the way to approach the topic of redundancy, or is there a better way? Any population geneticists out there with ideas?
Any thoughts would be greatly appreciated!
Sincerely,
John Bracht
[ 01 August 2002, 02:12: Message edited by: John Bracht ]
IP: Logged
|
|
charlie d.
Member
Member # 159
|
posted 01. August 2002 12:44
John:
I am not sure there are any known entirely and unequivocally redundant systems in biology. The main problem being, of course, that it is impossible to test for complete functional overlap, unless one knows exactly all the potential functions of a system, in all potential selective conditions an organism can find itself. For instance, some knock-out mice that initially were thought to have "no phenotype" were later discovered to have subtle but detectable alterations in specific functions (in my field, IgD-deficient mice are a classic example: see here and then here).
One other problem is that primary sequence may only be partly informative as far as functional diversification goes, as duplicated genes may differ in their pattern of expression, or may be expressed in rate-limiting amounts.
One recent paper that may be of interest to you is this one, about a database of duplicated genes in teleosts (which have likely undergone multiple genome duplications). Should be a good source of material.
[ 01 August 2002, 12:45: Message edited by: charlie d. ]
IP: Logged
|
|
James A. Barham
Member
Member # 50
|
posted 01. August 2002 14:55
This is more in the nature of a question than a comment (or maybe a little of both).
How would an engineer define a "redundant" system in such a way that it could be distinguished from a "robust" system?
It seems to me from reading thebiological literature (admittedly, with a layman's understanding) that there is some ambiguity in the use of these terms. I don't think they should just be interchangeable. Yet, one sometimes reads of "massive redundancy" in a system---like the brain, for instance---where what is intended seems to be something more like "robustness," in the sense of smooth degradability to failure and insensitivity to knockout of some or even many components.
What I suspect---but do not feel secure enough to assert---is that "redundancy" in the strict sense of a back-up system identical to the main system, is a purely mechanistic concept that will find no counterpart in organisms. But I could be wrong, of course.
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 01. August 2002 17:12
John,
To begin, redundancy is an almost universal, and essential feature of biological systems. Almost all biological systems have extra cells providing some degree of excess capacity or redundancy. In addition most biological systems have the ability to generate additional capacity if required or if some is lost due to accident. For the most part, biological systems provide for redundancy in a very efficient manner.
Since in sexual reproduction the vast majority of selection occurs prior to conception, the evolvability of any feature impacting survival after conception is problematic(using current theories and models). I can’t off hand see why redundancy is a particular problem, but I haven’t given it as much thought as you have.
IP: Logged
|
|
fish
Member
Member # 213
|
posted 01. August 2002 20:09
Hello John,
I think that there is a reasonably large population genetics literature out there on the evolution of redundancy.
I think it started approximately with Fisher who wrote a paper on the evolution of dominance.
As I recall he claimed that most harmful mutations were recessive because there had been selection to make things this way.
Wright and others refused to believe him and that particular idea has been pretty much refuted, largely because seletion is too weak at any given trait.
We are redundant to the extent that we have two copies of every gene. That is quite a lot of redundancy!
I think in general that one has to be a little careful. It is difficult to prove or disprove phenotypic redundancy in general. There are many knock outs that you can do that have no observable effect. But just because we cant see it doesnt mean that it doesnt have a significant effect in the wild.
It is correct to say that things selection does not see are redundant and will be removed by genetic drift. But genetic drift happens pretty slowly, so in many cases small countervaling factors can be enough to preserve the redundancy.
One interesting case are phase variation genes in bacterial pathogens. Bacteria like Neisseria meningitidis have more copies of certain surface proteins than they need at any one time. Some of them are turned off. But because human populations gain an immune response to specific variation over time, selection regularly changes
and a different protein becomes better than the predominant one in the population.
In this case, there is selection for redundancy in that bacteria with currently unused proteins have a better long-term evolutionary future than those without. In this case, the selection is only effective because the rate of loss of redundancy through genetic drift/selection for smaller genome size is sufficiently low. Otherwise it would disappear, however beneficial it was.
IP: Logged
|
|
Ryan Huxley
Member
Member # 366
|
posted 01. August 2002 21:35
This will be a minimal post on my part due to time constraints; though I hope to post more on this particular topic since I'm at least familiar with it from a structural engineering perspective.
To respond to what was noted by charlie d:
"I am not sure there are any known entirely and unequivocally redundant systems in biology. The main problem being, of course, that it is impossible to test for complete functional overlap, unless one knows exactly all the potential functions of a system, in all potential selective conditions an organism can find itself. "
At this point, I don't think it is necessary to determine ALL aspects of redundancy for given redundant system. I think that starting off with just one aspect of a given system that is redundant with another should be tried first. After these "simpler" examples of redundancy are understood, then the more complicated and all-inclusive ones can be investigated. So, all we need to do is just find a simple redundancy in one given area/system and focus on that one.
To respond to a query by James Barham:
"How would an engineer define a "redundant" system in such a way that it could be distinguished from a "robust" system?"
I think what you are probably referring to is more along the lines of redundancy vs reliability. In other words, redundancy helps to promote reliability - which, I suppose, is similar to the way you appear to be using "robustness." There are some mathematical equations that can be used to determine the reliability that a redundant system (say, a braced frame) confers for an "overall" system (say, a building). Trying to apply this mathematical expression to biological entities may be problematic, but it is something I've been attempting to do recently, with some helpful suggestions from John Bracht.
To respond to something "fish" has noted:
"I think that there is a reasonably large population genetics literature out there on the evolution of redundancy."
I would be interested if you could provide any references for that. While I am aware of studies of redundancy and "anti-redundancy," I have not yet come across studies for the origins of redundancy. While that is probably the area I am most interested in, other areas that provide examples of redundancy sustainability would also be appreciated since it seems that redundancy is, as John mentioned, rather expensive to maintain for limited, if any, advantages provided to an organism given their quiescent nature.
IP: Logged
|
|
charlie d.
Member
Member # 159
|
posted 01. August 2002 23:04
Regarding fish's comment, I am not sure that just because we have 2 copies of many genes for which a copy is sufficient, the other copy can be called redundant: there are very strong selective pressures for diploidy per se, some of which have nothing to do directly with gene function, but with the evolutionary advantages of sexual reproduction and genetic recombination. Alas, you can't be diploid for only some of your genes! That's however a good illustration of the problem I am talking about: sometimes apparent redundancy just means you are not looking at the right phenotype, or at the right level. Sometimes, it's just a result of other constraints (perhaps we could do at least equally well with one, maybe slightly larger kidney, but we are bilateral animals, and 2 don't seem to be enough of a problem).
Another example: you can knock out one copy of the Rb tumor suppressor gene from a cell and conclude that 2 copies are redundant for cell cycle checkpoint regulation - the cell does just fine thank you - however, children bearing a mutation of one of their Rb loci almost invariably develop retinoblastoma (the reason being that secondary mutations of the normal allele are not uncommon, and cells lacking both alleles very easily become cancerous: what is redundant at the cell level, is not at the organism's).
This also bears for Ryan's comment: I'd say you must consider all potential functions of a putative redundant system before you can even claim it is redundant in the first place. [A clarification: I assume here we are discussing the kind of redundancy John is talking about, which can hardly be explained by darwinian mechanisms]
[ 02 August 2002, 09:18: Message edited by: charlie d. ]
IP: Logged
|
|
James A. Barham
Member
Member # 50
|
posted 02. August 2002 08:56
Warren alludes to the capacity of organisms for regeneration and repair. This is perhaps a better avenue for getting at the question I was asking.
That question, again, was how to differentiate between "redundancy" in the strict sense of a "back-up" system vs. "robustness" in the sense that all biological systems have a general quality of homeodynamic stability or adaptability to varied circumstances.
So, consider Warren's point: my body can repair cuts in my skin and breaks in my bones; a salamander can grow new limbs; a dog can learn to walk successfully with only three limbs; a small piece of an amoeba or a hydra can grow into a complete organism. Do these capabilities have anything to do with "redundancy" in the literal, engineering sense? I wouldn't think so. If not, then why call other examples that have been cited "redundant"? Why not just acknowledge that all such capabilities are similar---part of the general ability of life to adapt to circumstances (within limits, of course)?
Here is another example to bring out my original point more clearly: it is widely assumed that brains operate according to principles analogous to artificial "neural nets," where "information" is stored in the "energy landscape" of the network as a whole, not in any single location. This gives brains great robustness, in the sense that many individual neurons can be knocked out with little or no degradation of functionality. This is similar in some respects to the phenomenon of holography (see the work of Karl Pribram). This obviously has nothing to do with "redundancy" in any usual sense of the term.
So, my question is, should the examples that are cited of "redundancy" be accepted as such on the assumption that the organism is just a kind of "mechanism," or should they not rather be viewed in the light of the general robustness (as exemplified in brains) that seems to be characteristic of all living systems?
IP: Logged
|
|
charlie d.
Member
Member # 159
|
posted 02. August 2002 09:41
I guess maybe we should distinguish between "hard" and "soft" redundancy.
"Soft" redundancy is akin to the "excess capacity" warren is talking about. This is hardly a problem for darwinian mechanisms: excess capacity for a system, generated by accidents or constraints dictated by evolutionary history, will be counterselected only if it generates an obvious disadvantage. We may all be carrying an extra kidney , or a useless pinky toe (just examples, I am not sure of either)- not because we necessarily need them, but because they don't hurt us.
"Hard" redundancy is the exact and complete duplication of a function (through a duplication of its components, or different ones), as a fail-safe mechanism. This may be necessary in some cases in which failures of the primary system is common (for instance, there is an obvious selective advantage to have both a DNA repair mechanism and a cell-cycle checkpoint system: DNA repair is not 100% efficient). On the other hand, there may be some that are not. I am thinking of the kind of multiple overlayed China syndrome fail-safe mechanisms that nuclear plants have: they are complex and exceedingly costly, and are likely to be never utilized, but still there they are, a testament to design teleology. These are actually quite common in human design, but I cannot think of any in biological systems. If they existed, these would be problematic for darwinian evolution. However, to convincingly prove they exist may be tough, for the reasons I described in previous posts.
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 02. August 2002 11:30
There are three issues here. One is the existence of redundancy in biological systems. The second is the evolvability of redundancy. The third is the impact/interpretation of shielding on evolutionary processes. A couple of additional comments on each of the issues.
Redundancy in a mechanical system designed by man refers in part to a back-up system or back-up capacity to address the possibility that the primary system or capacity is harmed or inadequate. In most practical applications, redundancy is addressed 1) by safety margins to prevent break-downs, 2)by providing spare parts and mechanics to fix broken components, or 3)by providing excess parallel capacity. Only under very specialized conditions would one ever engineer a system with an independent back up system.
The suggestion that redundancy in biological systems is difficult to identify and analyze does not seem consistent with the facts. Biological systems exhibit all the forms of redundancy used in systems designed by man( and probably do it better). [In response to James, the examples you list are examples of redundancy. ]
It is also not entirely accurate to suggest it is difficult to analyze/test the ‘evolvability of redundancy’. Consider as a possible example, James’ example of skin repair. Knowing what we do of redundancy in man made systems, we can identify in some detail the set of operations required to provide the skin repair mechanism. Starting with this list of operations we could obtain a rough estimate of the minimum number of genes and mutations required to produce and/or refine the repair mechanisms. Using what is known of mutation and selection rates it would not be difficult to create detailed ‘could have happened’ models of the evolvability of redundancy.
Using observations of differences in redundancy processes and mechanisms in related species, it might actually be possible to validate(or invalidate) the assumptions used in a ‘it could happen that way’ model. It almost certainly would be relatively easy to construct models to simulate the evolution of known examples of ‘changes in redundancy’. It is also, IMO, almost certain that the assumptions used in the model will not validate. It will almost certainly be found that the evolvability of redundancy in nature is faster, more precise, and involves fewer genetic changes, then any simulation based on Darwinian concepts. [Charlie’s comment that Darwinian evolution has no problem with ‘soft redundancy’ is something that could be tested/demonstrated with simulation models. No one has, to my knowledge produced, such a simulation. I would be to discuss in more detail the basis for my opinion that such tests/demonstrations would be problematic. At this point, any comment on the ability of Darwinian concepts to explain or not explain the evolution of ‘soft redundancy are pure unsupported speculation(IMO). ]
The main point raised in John’s original post was not, however, redundancy or the evolution of redundancy in the sense discussed above, but the more general question- "Does evolution theory work when the evolutionary change is only very slightly beneficial?" In other words, does what is called ‘shielding’ make evolutionary change less likely or more difficult. As I briefly alluded to in my original note, a look at shielding leads to some interesting observations on the role, purpose, or function in nature of dying, reproducing, and natural selection.
Darwinian theory defines the function of dying, reproducing and natural selection in terms of producing ‘selecting the fittest’ or generating evolutionary or adaptive change. In somewhat simplistic terms, the more a trait is exposed to differential survival and reproduction the faster it should evolve. The more a trait is shielded from differential survival and differential reproduction the slower it should evolve.
If we look at sexual reproduction in complex organism we find an interesting phenomena. Something greater than 99.999% of natural selection or selective survival and reproduction occurs before the male and female sex cells combine and before most phenotype traits appear. Most traits in nature, it would appear, are heavily shielded from Natural selection.
The explanation, I would suggest, is that natural selection or ‘dying and reproducing’ serve a purpose or function in nature much different from that suggested by Darwin. Despite a high degree of redundancy, cell division creates distortions(mutations) in the genetic code. By the end of an organisms life time, malfunctioning cells (cancers) are not uncommon. It seems likely that the primary function or purpose of dying and reproduction is not selection, but gene repair or maintaining the integrity of genetic code. The extensive selection occurring prior to sex cells combine, suggest that death and differential reproduction occur not to produce evolutionary change, but primarily to reverse the deleterious impact of accumulative mutations. [Variations of this interpretation, I have been told, have been proposed by number of different authors. ]
IP: Logged
|
|
charlie d.
Member
Member # 159
|
posted 02. August 2002 12:42
quote:
Charlie’s comment that Darwinian evolution has no problem with ‘soft redundancy’ is something that could be tested/demonstrated with simulation models. No one has, to my knowledge produced, such a simulation. I would be to discuss in more detail the basis for my opinion that such tests/demonstrations would be problematic. At this point, any comment on the ability of Darwinian concepts to explain or not explain the evolution of ‘soft redundancy are pure unsupported speculation(IMO).
I am wary of "simulations" of biological phenomena, when you can study the real thing. If a redundant trait has an obvious, empirically detectable selective advantage to the organism, it is up to the doubters to provide evidence against a role of natural selection in its origin, or in favor of a more parsimonious mechanism.
quote:
If we look at sexual reproduction in complex organism we find an interesting phenomena. Something greater than 99.999% of natural selection or selective survival and reproduction occurs before the male and female sex cells combine and before most phenotype traits appear. Most traits in nature, it would appear, are heavily shielded from Natural selection.
Can you please give some specific example? I truly can't imagine what traits you are thinking of, and you are claiming it's the majority of them.
[ 02 August 2002, 13:05: Message edited by: charlie d. ]
IP: Logged
|
|
fish
Member
Member # 213
|
posted 02. August 2002 19:13
Hello folks.
Ive been asked a worthwhile but time consuming question - provide some literature on the evolution of redundancy.
I must apologise that I am unlikely to do this in a truly conscientious fashion anytime soon. Fisher published his paper on the evolution of dominance in 1934 or something. So one way to go about it would be to look for papers that cite that...
I also think there was a paper by Maynard Smith and Martin Novak in Nature a few years back that was quite relevant.
Just found it:
125: Nowak MA, Boerlijst MC, Cooke J, Smith JM. Related Articles
Evolution of genetic redundancy.
Nature. 1997 Jul 10;388(6638):167-71.
PMID: 9217155 [PubMed - indexed for MEDLINE]
There is a paper in PNAS this year that might also be a good starting point...
Redundancy, antiredundancy, and the robustness of genomes.
Krakauer DC, Plotkin JB.
Santa Fe Institute, 1399 Hyde Park Road, Santa Fe, NM 87501, USA. krakauer@santefe.edu
And this:
: Dover G. Related Articles
How genomic and developmental dynamics affect evolutionary processes.
Bioessays. 2000 Dec;22(12):1153-9. Review.
PMID: 11084631 [PubMed - indexed for MEDLINE]
or this:: Krakauer DC, Nowak MA. Related Articles
Evolutionary preservation of redundant duplicated genes.
Semin Cell Dev Biol. 1999 Oct;10(5):555-9. Review.
PMID: 10597640 [PubMed - indexed for MEDLINE]
or this from a different perspective looks quite
interesting:
Tononi G, Sporns O, Edelman GM. Related Articles, Free in PMC
Measures of degeneracy and redundancy in biological networks.
Proc Natl Acad Sci U S A. 1999 Mar 16;96(6):3257-62.
PMID: 10077671 [PubMed - indexed for MEDLINE]
This would probably be a good place to look for the classical population genetics literature:
: Ohta T. Related Articles
Time for spreading of compensatory mutations under gene duplication.
Genetics. 1989 Nov;123(3):579-84.
Anyway, there is definitely more where that came from . For example, just look in PUBMED for Evolution AND redundancy.
[ 02 August 2002, 19:20: Message edited by: fish ]
IP: Logged
|
|
Jules
Member
Member # 181
|
posted 05. August 2002 18:15
Just a question or two. Regarding "junk" DNA, I've read somewhere that some of it was thought to code for redundant systems. How does Darwinian evolution explain how non-coding DNA is preserved? Is there a way to discover whether it really codes for something?
IP: Logged
|
|
|
Printer-friendly view of this topic