More Recent Comments

Monday, January 20, 2014

Can some genomes evolve more slowly than others?

I've been teaching my students about random genetic drift, phylogenetic trees, and the molecular clock. It's hard for undergraduates to understand that trees based on sequences are reflections of the fixation of nearly neutral alleles by random genetic drift. That's because they, like almost everyone else, think of evolution in terms of natural selection and adaptation.

It's even harder to grasp the idea of a molecular clock even though it's been around for fifty years. It was back in the 1960s that scientists like Emanual Margoliash noted that the rate of substitution of amino acids in every lineage was remarkably similar [The Modern Molecular Clock]. We now know that this is because the alleles are fixed by random genetic drift and that the rate of fixation by drift depends only on the mutation rate. It looks like the mutation rate is relatively constant in all lineages (bacteria, protozoa, plants, animals, etc.). This isn't a big shock since the vast majority of mutations are due to errors in DNA replication and the fundamental biochemistry of DNA replication and repair are similar in all species.

Calibrating the molecular clock involves relating the rate of substitutions to the number of years that have passed since two lineages diverged. This is a lot more difficult than it seems because it requires fossils that lie close to the node and accurate dates. Dan Graur and Bill Martin have likened this to Reading the entrails of chickens. But you don't need to calibrate the molecular clock in order to show that it exists.

This brings me to a paper that's just been published in Nature. The authors sequenced a cartilaginous fish, the elephant shark (Callorhimchus milii). This is interesting because the cartilaginous fish (Chrondrichthyes) and the bony vertebrates (Osteichthyse) are thought to have diverged about 450 million years ago (Myr). This is the first complete genome of a cartilaginous fish.

The genome is only about 1/3 the size of the human genome and it has about 19.000 protein-coding genes and hundreds of other genes that specify various RNAs. The authors constructed phylogenetic trees using a set of 699 genes that had orthologues in 12 other chordates. They confirmed that the cartilaginous fish (sharks) diverged early from the bony vertebrates, as expected.

But the big news—see the cover of Nature—is that the genome of the elephant shark is the slowest evolving vertebrate genome. Here's how the authors describe their result ..
Previous studies based on a few mitochondrial and nuclear protein-coding genes indicated that the nucleotide substitution rate in elasmobranchs is an order of magnitude lower than that in mammals16, 17. Using the genome-wide set of 699 orthologues, we estimated the molecular evolutionary rate of C. milii and compared it with other gnathostomes, with sea lamprey as the outgroup. Callorhinchus milii protein-coding genes have evolved significantly slower than all other vertebrates examined (P < 0.01 for all comparisons; Supplementary Tables VI.1–VI.3), including the coelacanth, which has been considered to be the slowest evolving bony vertebrate.

You have to look at page 181 (!!!) in the supplementary information in order to see the data. There you'll find a table showing relative rate tests. This is the standard way of testing for a molecular clock. What you do is compare the number of changes in two different lineages using an outgroup to root the branches. In this case, the authors used lamprey as the outgroup and compared the number of changes along the shark branch to the changes along branches leading to human, mouse, cow opossum, chicken, lizard, Xenopus, coelacanth, stickleback, and zebrafish.

The human/shark relative rate test is shown on the left. There were 15,046 amino acid changes in the lineage leading to modern humans and 14,154 along the branch leading to the elephant shark. Thus, the rate of evolution of sharks is only 94% of the rate for the human genes. The shark sequences showed the lowest number of substitutions in every single rate test. The lowest was 85% for coelacanth.

These substitutions are almost certainly neutral substitutions but the results could be confounded by selected substitutions (adaptations) if there were a significant number of them. The authors of the paper decided to look at nucleotide substitutions at fourfold degenerate sites in coding regions. This is one way to focus on sites that are presumably neutral. They say ...
A neutral tree based on fourfold-degenerate sites indicated that the low evolutionary rate is a reflection of the neutral nucleotide mutation rate, and confirmed that the neutral evolutionary rate of C. milii is the lowest (Fig. 2a).
The difference here is larger than that determined by the relative rate tests but it's still nowhere near the "order of magnitude" difference reported in previous publications.

It looks like there are about 10% fewer changes that have been fixed in the elephant shark genome compared to many other vertebrates. This doesn't seem like a big number to me since we're looking at stochastic changes over hundreds of millions of years. I prefer to see this glass as half full—there is an approximate molecular clock. I also remain a bit skeptical of the results since there are many potential sources of error.

But what if the data actually reflects a true slowing down of evolution in elephant sharks? What does this mean?
The whole-genome analysis of C. milii, a holocephalan cartilaginous fish, shows that the C. milii genome is evolving significantly slower than other vertebrates, including the coelacanth, which is considered a ‘living fossil’. Although several physiological and environmental factors have been proposed to explain the interspecific variation in molecular evolutionary rates, the factors contributing to the lower evolutionary rate of C. milii are not known.
The simplest explanation is that the biochemical mutation rate in elephant sharks is lower than in other species. In other words, DNA replication is more accurate in sharks or repair is more efficient. While we can't rule this out, it doesn't seem very likely.

Perhaps the explanation is much more complicated. Michael Lynch has some good arguments for a correlation between genome size and overall per nucleotide mutation rate per generation. Species with larger genomes tend to have larger mutation rates (Lynch, 2007, 2010). Note that the elephant shark genome is only about 1 Gb whereas mammalian genomes are about 3 Gb in size.


Lynch, M. (2007) The origins of genome architecture. Sinauer Associates, Inc., Sunderland, Massachusetts, USA

Lynch, M. (2010) Evolution of the mutation rate. Trends in Genetics 26:345-352. [doi: 10.1016/j.tig.2010.05.003]

Venkatesh, B. et al. (2014) Elephant shark genome provides unique insights into gnathostome evolution. Nature 505:174–179. [doi: 10.1038/nature12826]

83 comments :

John Harshman said...

It seems to me that there are quite a few potential variables -- e.g. repair efficiency, frequency of germ line replication, frequency of transcription, etc. -- affecting the rate of neutral evolution, so we shouldn't be surprised to see no single, unified clock. And the literature is full of evolutionary rate variation both within loci among taxa and within taxa among loci. Just one simple example: tinamous, as a group, are evolving about twice as fast as other paleognaths in all of the 20 loci I've looked at. A few minutes' work could summon up another dozen examples just within birds.

Joe Felsenstein said...

The stochastic models used in inferring phylogenies assume a (stochastically) constant rate of substitution in each branch of the tree. Branch lengths may or may not be proportional to time.

I'd add one more to John Harshman's good list. Substitutions may also be the result of occurrences of natural selection. If these occur sporadically and appear to randomly affect different sites, natural selection could come close to satisfying a stochastic substitution model. (Except for the resulting organism's fitness).

So it may be hard to prove that substitutions are neutral if selection appears random. (Note I am saying "appears", to avoid tedious arguments about determinism, the Copenhagen Interpretation of Quantum Mechaniscs, etc.)

Devin said...

There's a newer paper by Michael Lynch where he takes effective population size into account, and it seems to account for a lot of the variation in mutation rate. http://www.ncbi.nlm.nih.gov/pubmed/23077252

I think it's a compelling argument for why mutation rate varies between different species.

Larry Moran said...

The most striking result is that mutation rates DON'T seem to vary between species. We're focusing on the exceptions and ignoring the rule.

Georgi Marinov said...

What do you mean by "don't seem to vary between species"? They vary by orders of magnitude. Perhaps you meant "between closely related species"?

P.S. It is not as simple as just $N_e$, what also matters is how much damage an elevated mutation rate does, which determines the selection coefficients for mutator alleles (then $N_e$ comes into play to set the thresholds beyond which you can't go by selection). Thus the difference between Figures 1B and 1C...

Larry Moran said...

Doesn't the relative rate test just look at differences? Those branch lenghts reflect change over time.

Yes, it's true that a confirmed adapationist could make up a scenario where most of the substitutions could be due to natural selection that just happens to operate with the same speed in lineages that inhabit drastically different environments (e.g. birds and zebrafish). If you believe that's likely, I'd like to talk to you about some homeopathic remedies that I can sell you. :-)

We've got plenty of evidence that the substituions we see in well-studied proteins are neutral. That's why bacterial enzymes can replace yeast enzymes, for example, even though they differ at many sites.

Joe Felsenstein said...

Larry, I'll be happy to buy those homeopathic remedies from you. Paying for them, of course, according to their dry weight.

Unknown said...

Forgive the intrusion of a naïf; but, just how constant and reliable are molecular clocks?

This article has left me scratching my head:
http://www.wired.com/wiredscience/2014/01/evolution-evolves-under-pressure/

Susan Rosenberg suggests that: “Cells actually decide to turn up their mutation rate when they are poorly adapted to the environment.”

I question whether higher mutation rates can themselves be considered an adaptive response to stress. The logic to me appears most circular.

Just the same – the constancy of molecular clocks as reliable evolutionary timepieces, would appear at face value, quite questionable.

Georgi Marinov said...

The idea has been around for a long time, but I am ashamed to say that I am not up to date with the theoretical work in this area (if there is any, anyone who happens to know the subject well, please share you knowledge here). But based on first principles I could imagine something like that as possible in single-celled organisms, provided the selective advantages are sufficiently strong.

However, I have hard time seeing how that's going to work for a multicellular organism with very long generation times. But I might be wrong

Unknown said...

Georgi

My intuitions coincide with yours.

Begging yet another question – can inconstant mutation rates in multi-cellular organisms be more apparent in some regions of their genomes and less apparent in other regions? My reading has it that some molecular clocks (specifically miRNA) can be remarkably constant and reliable compared to others.

I am a big fan of Kevin Peterson’s work on miRNA phylogenetics
http://mbe.oxfordjournals.org/content/early/2013/08/02/molbev.mst133

Sadly, I am ill equipped to objectively and critically evaluate the validity of his team’s calibrations.

John Harshman said...

I'm willing to believe that most substitutions we see are neutral. But most substitutions we don't see are deleterious, and what's neutral clearly changes depending on what substitutions have already happened. If we imagine a sequence x fitness space for the protein (and I know you love your fitness spaces), it must have a very large but lacy peak consisting of lots of points connected by very thin lines with a lot of valleys taking up most of the space. Proteins certainly don't evolve at a neutral rate.

John Harshman said...

Well, fixation rates vary by orders of magnitude. We're just assuming, in all but a very few cases, that this is a reflection of mutation rates. (Though I think it's a correct assumption.)

AllanMiller said...

Georgi,

However, I have hard time seeing how that's going to work for a multicellular organism with very long generation times..

Another issue for eukaryotes in general is the predominance of recombination. If a mutator succeeded, and shook out a useful adaptation, it would have to be on the same chromosome for one, and the closer the better. The stress and benefit would need to happen with a sufficient frequency to maintain the mutator in the population. And it would tend to shoot itself in the foot, by mutating its own sequence along with the rest.

Alex SL said...

trees based on sequences are reflections of the fixation of nearly neutral alleles by random genetic drift

I do not understand this sentence. If I sequence only one individual per species, as is customary in many phylogenetic studies, where does the fixation enter? It may well be possible that lineage sorting has not yet taken place. We simply cannot see whether it has without much deeper sampling.

From what I have seen it is pretty well established that rates of substitution differ widely, especially when lineages evolve to be very short-lived, very long-lived, or parasitic.

As a side remark, page 181 of the supplementary data? This is crazy stuff.

Federico Abascal said...

Thanks for the post and discussion.

I would like to add something else that should also be considered. If gene duplications happen, then evolutionary rates of paralogs (orthologs with respect to a different species) will probably vary since the evolutionary pressure on these genes will not the same as was before. The story can be further complicated if independent gene losses occur. What we can learn from this is that the selective pressure on a gene is not necessarily the same in different species, and depends on: changes in other genes, environment, adaptation, speciation events, and so many other things already cited.

Joe Felsenstein said...

As most trees are still one individual per species, Alex SL's point is correct. But if the divergence times of the closest species are great enough, most of those differences probably will be fixations.

Let me also (1) note the phrase "nearly neutral". Nearly neutral mutations do not show the same dynamics as neutral ones (2) back strongly John Harshman's excellent point about the shifting role of selection against deleterious alleles. That was the original point of Fitch and Markowitz's 1970 paper about "covarions" -- that which substitutions are allowed may change along the tree.

Let me also add a comment on clocks. The molecular clock is universally used for work within species. It is also fairly reliable for groups of closely related species. As species become more and more divergent, changes in the rate of molecular evolution occur owing to their altered biology. So the molecular clock is neither universal nor totally useless. It is instead a useful and fruitful approximation.

You'd be crazy to use a clock on a tree that has both Human and Yeast in it. You'd be crazy not to use it on a tree that has only closely-related sparrows on it.

I have somewhat lost patience with arguments for/against the clock that draw one extreme conclusion or the other.

Larry Moran said...

John Harchman says,

Proteins certainly don't evolve at a neutral rate.

Yes, they do ... to a first approximation.

Larry Moran said...

Federico Abascal says,

What we can learn from this is that the selective pressure on a gene is not necessarily the same in different species, and depends on: changes in other genes, environment, adaptation, speciation events, and so many other things already cited.

Actually, what we learn from molecular evolution is that there is very little selective pressure on most genes at most times. That's why >99% of the changes we see have nothing to do with natural selection.

NickM said...

I think he means if you compared the DNA substitution rate in noncoding junk DNA to the DNA substitution rate in protein-coding DNA, the latter would be slower, reflecting stabilizing selection against deleterious mutations. But if the authors are looking at fourfold degenerate sites in protein-coding regions, those sites should be close to neutral.

I worry about the statistical issue of those long branches without close relatives, you will detect more substitutions when you add more branches. Did they run the tests to check for this?

Anonymous said...

A population bottleneck can look like accelerated mutation rates because the small population size results in a higher proportion of fixed neutral and semineutral alleles (and, yes, some not-too-good ones).

Larry Moran said...

Georgi Marinov says,

[Mutation rates] vary by orders of magnitude.

I don't think this is correct unless you are counting viruses as species. The mutation rate is pretty constant as far as we know. There are populations that have higher mutation rates because of mutations in repair genes but that's temporary. It's not a characteristic of the species.

Can you give me an example of two species whose mutation rates differ by more than 100-fold (i.e. two order of magnitude)?

Larry Moran said...

Joe Felsenstein says,

You'd be crazy to use a clock on a tree that has both Human and Yeast in it.

Why do you say that? I agree that the accuracy of molecular clock calibrations decreases the further back you go but you can still get some reasonable estimates.

The molecular clock data suggests that multicellular animal phyla diverged about 700 million years ago. I think the split between yeast and animals is about twice as deep in most trees. Therefore, it seems reasonable to conclude that fungi and animals diverged about 1.4 billion years ago. What's wrong with that? Why would I be "crazy" to reach such a tentative conclusion?

I have somewhat lost patience with arguments for/against the clock that draw one extreme conclusion or the other.

Do you think it was an "extreme conclusion" to say that elephant sharks are the slowest evolving vertebrate?

John Harshman said...

Two problems here. First, there's a big space between "pretty constant" and "two orders of magnitude". Would you consider factors of two or three to be "pretty constant"? Second, are you willing to use substitution rate as a proxy for the mostly unknown mutation rate?

John Harshman said...

Nick, I think Larry is talking about amino acid substitutions, not fourfold degenerate sites. I have no idea how he can possibly think those evolve at a neutral rate. Any check of any exon data set should reveal the huge discrepancy between 2nd positions and 3rd positions.

And I agree with your point about long, unbroken branches being likely to underestimate true change.

Unknown said...

I've found this on youtube:
http://www.youtube.com/watch?v=Za1nbINxdP8

"Impact of Research on Race Formation and Mutations on the Theory of Evolution." Biologist Maciej Giertych speaks on the various scientific problems with Darwinian evolution."

Presents the process of genetic drift as process in the opposite direction of evolution as loss of genes and loss of variation.

Where does he go wrong?

Faizal Ali said...

A 56 minute video by some creationist anti-Semite? You don't seriously expect anyone to watch that, do you? Why don't you provide a reference for some of his peer-reviewed research on the topic, instead?

Piotr Gąsiorowski said...

Maciej Giertych is the only Polish young Earth creationist of note. He's a bona fide dendrologist and has done some useful research on subjects connected with forestry, but he's completely ignorant of evolutionary theory, no doubt as a matter of personal choice (wilful rejection of knowledge by a brain infected with YEC memes).

He's wrong about a great many things. What he shows in the video is a poor man's version of population genetics, using slogans and demagoguery in lieu of maths.

For Giertych, mutations are practically always deleterious (WRONG) and inevitably lead to degeneration (WRONG), while selection and drift reduce the genetic diversity of populations (RIGHT) and leave the population "impoverished" (well, they do eliminate alleles, but it doesn't mean what Giertych thinks it means). Therefore, there is no way a novel advantage can result from (micro)evolution (NONSENSE), and all advantageous heritable traits must have been created by God a few thousand years ago; otherwise they would have decayed by now (YEC GOBBLEDYGOOK).

Wikipedia reports him as saying: I am a scientist — a population geneticist with a degree from Oxford University and a PhD from the University of Toronto — and I am critical of the theory of evolution as a scientist, with no religious connotation. Which is of course a blatant lie. He's a living demonstration that neither legitimate degrees nor a professorial tenure guarantee one's credibility as a scientist.

Piotr Gąsiorowski said...

Unknown, if you are interested in real population genetics (as opposed to its creationist travesty peddled by Giertych), Prof. Felsenstein has generously made the electronic version of his excellent book on Theoretical Evolutionary Genetics available online:

http://evolution.genetics.washington.edu/pgbook/pgbook.html

Georgi Marinov said...

Can you give me an example of two species whose mutation rates differ by more than 100-fold (i.e. two order of magnitude)?

http://www.pnas.org/content/109/45/18488/F1.large.jpg

John Harshman said...

There seems to be some ambiguity about how we're measuring rates here: per genome replication, per generation, or per year. Relative rate tests measure (relatively) the latter. Your paper is measuring per generation, I think, which is for most of the taxa involved, the same as per replication. But adjusting that would make Larry's case worse, not better.

Amy Blackwelder said...

In regards to the paper of genetic drift at this blog. I'm trying to understand genetic drift versus natural selection better.

I understand how an accident might kill off much of a population and thus yield a population with less variation and thus more homozygous…

But How does genetic drift cause a population to become homozygous just by chance with parents having offspring in the example of one isolated group becomes “A” and another isolated group becomes “a” It seems that chance would just as much yield the “a” as the “A” in the same isolated group? That natural selection would be what causes the group to completely loose either the “a” or “A”, because groups will choose that which is common to mate.

How is only one set of genes passed on by chance alone making an entire isolated or small population homozygous? This might be an issue of my lack of understanding of chance with genes.

The secondary part of my question:
The paper says "Since the only way for neutral mutations to become fixed in a population is through random genetic drift..."

My question- If the population is a small tribe and they revered an extra thumb per se
then the females would choose to mate with those males and keep the neutral gene in the gene pool...or am I missing something? Isn't that also natural selection? or how is that genetic drift?

Federico Abascal said...

Lawrence A. Moran said: "Actually, what we learn from molecular evolution is that there is very little selective pressure on most genes at most times. That's why >99% of the changes we see have nothing to do with natural selection. "

I would rephrase it in the opposite way: the changes that we do not see (conservation) are the ones that best reflect natural selection. The observed ones are most likely neutral or nearly neutral (I guess this is what you meant).

What I would say is that the strength of purifying selection depends on many factors and that, at the end, weaker or stronger pressures are reflected in the number of accumulated mutations.

Larry Moran said...

@Federoco Abascal,

Oops! Point taken. There is strong negative selective pressure that prevents change.

I was only thinking of the causes of evolutionary change.

Larry Moran said...

@Georgi,

Thanks for the link. That's a Lynch paper that I haven't read. Looks like you are right and I am wrong. There are species with a 100 fold difference in mutation rates. I should have remembered from Lynch's book.

If the data is correct then we should see 10-fold less change in Chlamydomonas lineages in all sequence trees. Unless, of course, the lower mutation rate is a relatively recent event.

Larry Moran said...

John Harshman asks,

Would you consider factors of two or three to be "pretty constant"?

Yes.

Second, are you willing to use substitution rate as a proxy for the mostly unknown mutation rate?

Yes, if we know that the substitutions are neutral.

John Harshman said...

I would not consider two or three to be pretty constant, and I doubt most biologists would. Only an astronomer would think so.

And of course we don't know that substitutions are neutral, we just think it likely.

Faizal Ali said...

You might find this helpful:

http://evolution.berkeley.edu/evosite/evo101/IIIDGeneticdrift.shtml

In fact, that whole Berkeley series is a good way to get a grasp of the bare basics of evolution.

John Harshman said...

In your example, that isn't genetic drift and that isn't a neutral mutation. The fact that females prefer extra-thumbed males makes that mutation highly advantageous.

As to how drift can cause neutral alleles to become fixed (which is what you mean by "homozygous" here), it doesn't happen all at once. Just by chance, the frequency changes a bit in each generation in any finite population. The expected frequency in the next generation is whatever it is in the previous generation, but there's a variance in the realized frequency that depends on the population size. This causes a random walk in allele frequencies, with frequencies going up and down at random, that will inevitably end, possibly after a very long time, with one allele reaching 100% frequency and the other reaching 0%. By the way, at any given time the probability that a given allele will eventually become fixed is equal to its current frequency.

Georgi Marinov said...

There were also these recent papers, which seem highly relevant to the subject of the post:

http://www.ploscompbiol.org/article/info%3Adoi%2F10.1371%2Fjournal.pcbi.1002785
http://gbe.oxfordjournals.org/content/5/7/1393.long

Amy Blackwelder said...

@John

That is what I don't understand though. If an extra thumb is not neutral because it is preferred for silly reasons, how does the silly reason make that gene advantageous? If the reality is that the extra thumb does nothing to help or hinder survival in nature?

Seems like the gene itself would be neutral.

Perhaps seeing the commonly refereed to list of neutral genes would help me picture this.

But my question wasn't that it was genetic drift, but that it was natural selection.




Amy Blackwelder said...

Your explanation helps a lot on the second question -if you can show an example of this ; however, that would be best. Thank you.

Amy Blackwelder said...

Lute- thanks for the website. I've actually looked at that before. I get that info.
What I don't get is the info I posted. It seems confusing to me.

Faizal Ali said...

Whether a gene is considered neutral or deleterious or advantageous is based entirely on whether its expression increases its owner's chances of successfully reproducing. It doesn't matter if the reason for this advantage seems "silly" to you.

"Survival" isn't the issue. Reproduction is. So, in your example, the fact that females find the extra thumb sexually attractive means those with extra thumbs are more likely to reproduce and propagate the gene that produces the extra thumb. Which means that gene if favoured by natural selection.

Faizal Ali said...

The classic example is the peacock's tail. Having such a cumbersome appendage, if anything, makes it harder to get around and evade predators. So on that basis alone it probably impedes "survival." But the chicks (or, more correctly, the hens) love it. So it's selected for.

Amy Blackwelder said...

So this process would also fall under natural selection then. Thx.

That doesn't actually clear up for me the question about what the gene would be labeled though...


Because Survival is an issue. It doesn't matter if you reproduce if your offspring don't survive.

So if this gene for an extra thumb was also linked to cause an illness...the gene would be chosen by the women -natural selection- but then the offspring would have a less chance to survive. (Not that the women or men would know this)

So the gene may eventually die off naturally in that group as well as that group.

So is this gene advantageous because it will be reproduced or deleterious because it will lead to an eventual end of survival?

With all other animals -but humans- the reasons for choosing a gene seem to be survival. Not other reasons. Animals are good at identifying 'problems with genes' and will even leave there young when they've sensed something is wrong with it.

Amy Blackwelder said...

Good example. But the reason the females like it is because the tail (though cumbersome for evading fast prey) also acts as a disguise and the tail itself you can see looks like big eyes. This confuses other animals into thinking it has a huge form, large than it has which also helps it to avoid many predators which otherwise may have wanted to attack it.

Unknown said...
This comment has been removed by the author.
Unknown said...

Again, forgive my intrusion as I am an aging and hopelessly out of date high school Biology teacher. I am very grateful for the opportunity to correct any naïve thinking on my part.

I teach my students that molecular clocks are little different than radioisotopes used in radiometric dating

http://anthro.palomar.edu/time/table_of_isotopes.htm

Depending on the timescale examined, different radioisotopes and different molecular clocks are required. For example Kevin Peterson makes a cogent case for miRNA molecular clocks to trace eumetazoan lineages in what could be called evolutionary somewhat "deep time"

http://www.sciencemag.org/content/334/6059/1091.abstract

You left me scratcing my head: why cannot rRNA be used as a "deeper time" clock to produce a tree that has both humans and yeasts on it. Perhaps the calibrations are relative and cannot pinpoint exact chronological points in time. But still, these trees must be providing some useful information or I am seriously missing something.

To summarize (for the benefit of my students)

Molecular clocks are not constant:

1. Different proteins within one organism can evolve at different rates.
2. Identical proteins in two different related species can evolve at considerably different rates (say between two closely related species of birds)
3. Different categories of genome evolve at different rates (say rRNA vs AA sequences)
4. Identical categories of genomes can evolve at different rates between lineages (The paper cited by Larry here)

All of which makes molecular clock calibration problematic but not necessarily impossible

I hope I got this all correct and am not missing anything.

Unknown said...

One last reply to Joe,

To follow up on what I want to say in the classroom:

Everything above all said:

Radioisotopes and molecular clocks different in one other very important aspect: Radioactive decay is very constant and unfluctuating; whereas molecular clocks can accelerate and decelerate (even within relatively short timeframes on an evolutionary timescale) resulting in data that may initially only provide relative time points until calibrated against the fossil record.

Unknown said...

Evolution is not synonymous to "Natural Selection". "Natural Selection is but one mechanism whereby Evolution can occur.

The scenario you are describing (mate-preference for the extra thumb) is an example of sexual selection, another mechanism for Evolution besides Genetic Drift which is yet another. http://evolution.berkeley.edu/evosite/evo101/IIIE3Sexualselection.shtml

Darwin was the first to elucidate that other mechanisms other than Natural Selection can drive evolution.
https://ebooks.cambridge.org/ebook.jsf?bid=CBO9780511703829

Amy Blackwelder said...

I get everything you said in the second part. So if the current frequency is 50% then the probability it will become fixed is 50%. So it is not always, ok, that makes sense.

But how is this true:
Since the only way for neutral mutations to become fixed in a population is through genetic drift.

Amy Blackwelder said...

My understanding is sexual selection is a mode of natural selection.
I realize there are mechanisms for evolution.
I also realize evolution is not synonymous to natural selection.

But you bring up a good point.

Amy Blackwelder said...


What I don’t get it how we can then say ALL neutral mutations are the result of genetic drift?

Another example: One green eye and one brown is called neutral -at least in my reading-

What if a culture loves the one green eye. Then they are choosing it directly. Natural selection. -Not genetic drift.

Does this gene then change its labeling from neutral to advantageous? And if so, how does this makes sense because the gene itself does nothing to help in survival.

I guess I'm just trying to fully understand these terms .

Georgi Marinov said...

What I don’t get it how we can then say ALL neutral mutations are the result of genetic drift?

We cannot say that because it is not true.

Drift acts on mutation, it does not cause them directly

Does this gene then change its labeling from neutral to advantageous?

We do not think of these things in binary terms, even if we sometimes speak of them in such a way. There is a selection coefficient associated with everything - and it's almost never exactly zero. What behaves "effectively neutrally" is determined by the population genetic environment. And the selection coefficient can, of course, change with changes in the environment.

Amy Blackwelder said...

Ok, everything you said makes perfect sense.

But where I get the statement is from here:
http://www.talkorigins.org/faqs/genetic-drift.html

I was agreeing with everything until he got to the last sentence.

"Studies of evolution at the molecular level have provided strong support for drift as a major mechanism of evolution. Observed mutations at the level of gene are mostly neutral and not subject to selection. One of the major controversies in evolutionary biology is the neutralist-selectionist debate over the importance of neutral mutations. Since the only way for neutral mutations to become fixed in a population is through genetic drift this controversy is actually over the relative importance of drift and natural selection."

Georgi Marinov said...

What you quoted and what I said are not mutually exclusive, in fact I don't even see an apparent contradiction.

There isn't much of a controversy though.

Amy Blackwelder said...

I'm not really concerned about the controversy.

I'm just trying to understand this one part

only way for neutral mutations to become fixed in a population is through genetic drift this

Amy Blackwelder said...

My point -which may very well be misinformed- was that it seemed to me that some neutral mutations being fixed was the result of natural selection...not genetic drift.

That most would surely be the result of genetic drift, since it is much stronger than natural selection (in that advantageous genes have a tough time remaining and disadvantageous genes have a statistically higher chance)

But some would be from natural selection.

Georgi Marinov said...

My point -which may very well be misinformed- was that it seemed to me that some neutral mutations being fixed was the result of natural selection...not genetic drift.

An even more underappreciated than drift phenomenon (though not one of the four evolutionary forces) is what people refer to as genetic draft - that's when you have a mutation that's under strong selection and because recombination is limited, other mutations that are in linkage disequillibrium with it get fixed too.

Is that what you were referring to by "neutral mutations fixed by selection"?

Amy Blackwelder said...

I hope this isn't convoluted. -but the crust of my question is the question of this sentence "that the only way for neutral mutations to become fixed in a population is through genetic drift"

But below it doesn't seem to be the case...

I don't know the technical terms, but it seems that some mutations which may have an advantage for one reason or another (be it survival or other) may be linked to genes that are neutral in a sense and therefore those neutral mutations survive, indirectly -catching the ride so to speak.

So, we might say the neutral mutation is a result of natural selection indirectly. not genetic drift. (or this is called genetic draft? -ok.) But still this isn't genetic drift? So, the sentence seems wrong.



Also, if mutations labelled as (neutral/advantageous/deleterious) can change their labeling based on environments (which I would say they do and it seems like that is what you said from above with the selection coefficient) than how can we really call anything neutral?

It would seem that we wouldn't really know if the "neutral" mutation at some point in the past was advantageous or disadvantageous and resulted from natural selection...simply because the environment we are observing it in is different from the one it lived in...

And therefore the fixed mutation in our body we are calling neutral currently could have survived because of some point in the past that made it advantageous. Not genetic drift.

And in the sense also that the human green eye/brown eye is in this environment called neutral, but if for some reason it became popular from a celebrity per se...and the mutation began to increase -then this is natural selection not genetic drift on a once neutral mutation.
(but this part would only happen in humans)

And now the mutation is no longer even called neutral, but advantageous for reproduction (but whether it helps the survival is a different subject altogether)

So, how can we define a mutation as neutral when the term can change depending on the environment and when the neutral mutations we have could have resulted from past pressures or selections? Perhaps many neutral genes we have today are the result of a disadvantaged or advantageous place in their past environment.

So it just seem possible to say all these such and such genes are neutral and they are in fact from genetic drift.

Maybe you have info that could help me on that

Georgi Marinov said...

Also, if mutations labelled as (neutral/advantageous/deleterious) can change their labeling based on environments (which I would say they do and it seems like that is what you said from above with the selection coefficient) than how can we really call anything neutral?

Because I said that a mutation may change from effectively neutral to being visible to selection, not that all mutations do.

Most mutations have no visible effect on the molecular level, let alone on phenotype. The vast majority of the genome is junk DNA - most mutations there have no effect on anything (unless they happen to create a new regulatory element for a nearby gene or something of the sort). Similarly for 4-fold degenerate sites in genes (setting "duons" aside), etc. And mutations that do change the sequence of proteins need not have fitness effect either, and many indeed don't.

John Harshman said...

Yes, it's the net effect that matters. A given allele may produce both advantageous effects (attractive to the ladies) and disadvantageous (causes illness). Fitness of an allele is the sum of all these various effects. And fitness of an individual is the sum of all the various effects of all the various alleles in his/her genome. Now, originally you had said that the extra thumb was neutral except for sexual selection, which clearly renders it advantageous overall.

Amy Blackwelder said...

Ok, that info helps me understand this more.

So, if the thumb was neutral in all survival aspects it would be called advantageous because of the net effect from the reproduction aspects.

And if the thumb carried an illness, it would be?

Amy Blackwelder said...

Yes I understood the may. Not all do. I get that.

But because we have this possibility (however small) it seems unfair to say ALL fixed neutral mutations are the result of genetic drift, which is what the above sentence implies.

With this mean that the one green eye -one brown eye is the direct result of genetic drift and nothing else because we call this mutation neutral in our current culture?

OK, I see what your saying on the second part. But does this really mean ALL?

The sentence : "that the only way for neutral mutations to become fixed in a population is through genetic drift"

One green eye is neutral to us now, but this gene a long time ago could have proved less useful to survive if in the dark it did not blend as well. So, it then would have been disadvantageous.

Perhaps my misunderstanding is with the word fixed. Does it mean a mutation that is set in the population or dominate in the population? I thought it just meant set as in the green eyed mutation.

Amy Blackwelder said...

And if later this extra thumb mutation is no longer considered favorable and females don't go out of their way to select it ...then the mutation would then be labeled neutral? (given that it has no effects other then an extra thumb)

Back to the original issue...

The sentence : "that the only way for neutral mutations to become fixed in a population is through genetic drift"

It doesn't seem to me that that is true in all cases. Possibly most cases, but not all.

That is all I'm saying.

Amy Blackwelder said...

For Georgi- Thank you for taking the time.
"Most mutations have no visible effect on the molecular level, let alone on phenotype."-

Yes, but what about the ones that do like the mutation for one green eye and one brown eye

Amy Blackwelder said...

Georgi- Oh, also-

I'd say that the least paragraph would be a mix of genetic drift and genetic draft, not just genetic drift.

Again- the original sentence from talk origins not being accurate. The sentence : "that the only way for neutral mutations to become fixed in a population is through genetic drift"



Most mutations have no visible effect on the molecular level, let alone on phenotype. The vast majority of the genome is junk DNA - most mutations there have no effect on anything (unless they happen to create a new regulatory element for a nearby gene or something of the sort). Similarly for 4-fold degenerate sites in genes (setting "duons" aside), etc. And mutations that do change the sequence of proteins need not have fitness effect either, and many indeed don't.

John Harshman said...

If the thumb carries an illness, as I said its fitness would be the sum of all its effects on reproductive success. To make it excruciatingly simple, if males with regular thumbs tend to have two healthy children while those with extra thumbs have four children, one of which dies of extra-thumb disease, the extra thumb is still ahead by one.

This, however, doesn't bear at all on the claim "that the only way for neutral mutations to become fixed in a population is through genetic drift", since the extra-thumb thing isn't neutral. It's always risky in biology to say there's only one way for something is happen. While drift is by far the most common way for neutral alleles to become fixed, Georgi Marinov has mentioned another one (though a typo may have rendered it unintelligible): genetic hitchhiking, in which a neutral allele very close in the genome to an advantageous allele becomes fixed just because the two alleles are inherited as a unit. I am currently unable to think of a third alternative.

John Harshman said...

Where did you get the idea that peacock tails work as disguises? Do you have a citation to the literature? Ditto for the claim that this is why female peafowl like the long tails.

John Harshman said...

Some biologists claim that sexual selection isn't a form of natural selection. Those biologists, however, are wrong.

Larry Moran said...

John Harshman says,

Some biologists claim that sexual selection isn't a form of natural selection. Those biologists, however, are wrong.

Finally, something we can agree on.

Now, let's try and go one step further. Do you agree that sexual selection is a very minor part of the overall history of life over the past three billion years? If so, can't we just ignore it? :-)

John Harshman said...

Thanks for the smiley. Of course, we could ignore it if we wanted to. But I will confess a bias in favor of organisms that are moderately close relatives of mine, like birds and insects, as opposed to all those boring eubacteria. Since sexual selection is important in the tiny little corner of life that I like to pay attention to, I won't ignore it at all.

I'm pretty sure we agree on a few other things, though.

John Harshman said...

Your problem seems to be mostly that you think sexual selection is not a form of natural selection. If having one green eye and one brown makes you more likely to find a mate, then this isn't a neutral character. It's advantageous. The preferences of prospective mates are part of the environment experienced by the individual. Survival is not the target of natural selection. Reproductive success is the target, and survival is only useful in so far as it contributes to reproductive success.

The problem with the talk-origins sentence is relatively minor: it fails to recall that some neutral alleles are fixed by genetic hitchhiking (sometimes called genetic draft). But hitchhiking accounts for a minuscule proportion of neutral fixations.

Georgi Marinov said...

For Georgi- Thank you for taking the time.
"Most mutations have no visible effect on the molecular level, let alone on phenotype."-

Yes, but what about the ones that do like the mutation for one green eye and one brown eye


What about them?

There is a selection coefficient $s$ associated with them, whether due to sexual selection, environmental factors, or some combination of those. Changes in $s$ (and/or $N_e$) determine what is most likely to happen with them. I really don't see what the issue is...

John Harshman said...

I see dollar signs around your s and N_e. What were they supposed to be (italics?) and how did that happen?

Georgi Marinov said...

Some blog platforms accept LaTeX formatting so things actually come out as math symbols, subscripts and other niceties like that. I keep trying in vain hope this one will start doing that too at some point, but it keeps disappointing me.

Amy Blackwelder said...

Your first paragraph clears that up. I can no understand it as a sum of all effects .That makes sense perfectly.

The second paragraph also is helpful to me, but which doesn't really clarify one way or the other---but that was my point (what Georgi brought up was one example). He named it genetic draft, but that is something I saw myself and so hesitated on the (sentence in question).
I also see other exceptions -which may or not be correct-


Amy Blackwelder said...

It was on a nature show.

Amy Blackwelder said...

@ John- That is not it;) I do think sexual selection is natural selection, it was Shay who said otherwise.

The reasons a neutral mutation is not just fixed by genetic drift (though most are):

Theoretically speaking-

1) Using just the example of the green eyed/brown eyed.

It was neutral at one point, not making a difference one way or the other in reproduction. Nor does it aid in survival.

A few people had it.

Now -theoretically- a lot of people like the gene for whatever reasons and so now it is advantageous.

Now lots of people get the gene.

Time passes. -theoretically- no one cares about this gene anymore. But much more of it is in the gene pool now.

Scientists study the genome and find this gene -which does affect phenotype. But at this time this gene is viewed as neutral because it doesn't affect anyone one way or another in reproduction or survival.

So- the problem is we are now calling it the result of genetic drift because it is a neural gene now ----but at one time (which we are not aware of) it was not.

This was one main point (But like @John said below we do have to be careful about saying anything is done only one way in biology.) I agree. Which is why the sentence bugged me.

But here we have an example of a mismatch of the labels (advantageous/neutral), because the gene actually has changed in its frequency because of selective pressures and natural selection- and in fact is not still found in the genome only because of genetic drift, but because at one time in the past it was advantageous.

2) The second main point was with Genetic Draft (as it is evidently called -learned the technical name by Georgi,) when I mentioned a mutation/gene taking a ride with an advantageous gene.

This is another way it is not genetic drift.

So, I guess the bottom line is that the sentence is mostly right, but not fully right...if only for reasons #2.

Amy Blackwelder said...

Further no one knows for certainty why the tail is so attractive, I tend to lean more toward the side of it is linked to their survival. When I looked at it, immediately I noticed all the eyes. For other animals I imagine a similar confusion is possible.

Cats love to raise their hair s and look bigger when scared, many animals try to appear bigger. Also some animals have false eyes for confusion like the butterfly, some fish,

(this multiple eye thing is actually common in nature)

It could also be to make predators think their head is somewhere it is not, so they have a better chance to escape.

Most predators key visually on eyes of their prey to avoid them to not be seen. If a predator is confused with where the eye actually is, then the prey has an advantage.


https://www.ebiomedia.com/why-do-some-animals-pretend-to-have-eyes.html

Faizal Ali said...

The more problematic part of your scenario is when you say "the reason the females like it is because the tail (though cumbersome for evading fast prey) also acts as a disguise and the tail itself you can see looks like big eyes" (my bold).

It may just be that you were careless with your wording, but as written your statement implies that peahens know that the trait of the tail being a disguise will be passed on to their offspring if they mate with a particular male. A view more consistent with evolutionary theory would be that the females' sexual preference for males with the tail camouflage will be positively selected for, as it increases the chance of their offspring surviving.

Amy Blackwelder said...

Well, yeah:) What I mean by this is that we are programmed to be attracted to that which survives. I don't mean the bird consciously is making these decision;)

Like I am attracted to physical characteristics in men that have proven survivable, even if I don't what these characteristics are or why.

John Harshman said...

I think the claim about peacocks is dubious and I doubt there's any real evidence for it. At any rate there are many sexually selected characters that have no obvious (even as obvious as this peacock idea) direct connection to survival advantages. Male whydahs, for example, have bizarrely long tails, and it's hard to say what advantage they could offer. A more likely idea is that these features are hard-to-counterfeit signals of some other good quality; that's the "handicap" theory.