More Recent Comments

Sunday, March 27, 2016

Georgi Marinov reviews two books on junk DNA

The December issue of Evolution: Education and Outreach has a review of two books on junk DNA. The reviewer is Georgi Marinov, a name that's familiar to Sandwalk readers. He is currently working with Michael Lynch at Indiana University in Bloomington, Indiana, USA. You can read the review at: A deeper confusion.

The books are ...
The Deeper Genome: Why there is more to the human genome than meets the eye, by John Parrington, (Oxford, United Kingdom: Oxford University Press), 2015. ISBN:978-0-19-968873-9.

Junk DNA: A Journey Through the Dark Matter of the Genome, by Nessa Carey, (New York, United States: Columbia University Press), 2015. ISBN:978-0-23-117084-0.
You really need to read the review for yourselves but here's a few teasers.
If taken uncritically, these texts can be expected to generate even more confusion in a field that already has a serious problem when it comes to communicating the best understanding of the science to the public.
Parrington claims that noncoding DNA was thought to be junk and Georgi replies,
However, no knowledgeable person has ever defended the position that 98 % of the human genome is useless. The 98 % figure corresponds to the fraction of it that lies outside of protein coding genes, but the existence of distal regulatory elements, as nicely narrated by the author himself, has been at this point in time known for four decades, and there have been numerous comparative genomics studies pointing to a several-fold larger than 2% fraction of the genome that is under selective constraint.
I agree. That's a position that I've been trying to advertise for several decades and it needs to be constantly reiterated since there are so many people who have fallen for the myth.

Georgi goes on to explain where Parringtons goes wrong about the ENCODE results. This critique is devastating, coming, as it does, from an author of the most relevant papers.1 My only complaint about the review is that George doesn't reveal his credentials. When he quotes from those papers—as he does many times—he should probably have mentioned that he is an author of those quotes.

Georgi goes on to explain four main arguments for junk DNA: genetic load, the C-value Paradox, transposons (selfish DNA), and modern evolutionary theory. I like this part since it's similar to the Five Things You Should Know if You Want to Participate in the Junk DNA Debate. The audience of this journal is teachers and this is important information that they need to know, and probably don't.

His critique of Nessa Carey's book is even more devastating. It begins with,
Still, despite a few unfortunate mistakes, The Deeper Genome is well written and gets many of its facts right, even if they are not interpreted properly. This is in stark contrast with Nessa Carey’s Junk DNA: A Journey Through the Dark Matter of the Genome. Nessa Carey has a PhD in virology and has in the past been a Senior Lecturer in Molecular Biology at Imperial College, London. However, Junk DNA is a book not written at an academic level but instead intended for very broad audience, with all the consequences that the danger of dumbing it down for such a purpose entails.
It gets worse. Nessa Carey claims that scientists used to think that all noncoding DNA was junk but recent discoveries have discredited that view. Georgi sets her straight with,
Of course, scientists have had a very good idea why so much of our DNA does not code for proteins, and they have had that understanding for decades, as outlined above. Only by completely ignoring all that knowledge could it have been possible to produce many of the chapters in the book. The following are referred to as junk DNA by Carey, with whole chapters dedicated to each of them (Table 3).


The inclusion of tRNAs and rRNAs in the list of “previously thought to be junk” DNA is particularly baffling given that they have featured prominently as critical components of the protein synthesis machinery in all sorts of basic high school biology textbooks for decades, not to mention the role that rRNAs and some of the other noncoding RNAs on that list play in many “RNA world” scenarios for the origin of life. How could something that has so often been postulated to predate the origin of DNA as the carrier of genetic information (Jeffares et al. 1998; Fox 2010) and that must have been of critical importance both before and after that be referred to as “junk”?
You would think that this is something that doesn't have to be explained to biology teachers but the evidence suggests otherwise. One of those teachers recently reviewed Nessa Carey's book very favorably in the journal The American Biology Teacher and another high school teacher reveals his confusion about the subject in the comments to my post [see Teaching about genomes using Nessa Carey's book: Junk DNA].

It's good that Georgi Marinov makes this point forcibly.

Now I'm going to leave you with an extended quote from Georgi Marinov's review. Coming from a young scientist, this is very potent and it needs to be widely disseminated. I agree 100%.
The reason why scientific results become so distorted on their way from scientists to the public can only be understood in the socioeconomic context in which science is done today. As almost everyone knows at this point, science has existed in a state of insufficient funding and ever increasing competition for limited resources (positions, funding, and the small number of publishing slots in top scientific journals) for a long time now. The best way to win that Darwinian race is to make a big, paradigm shifting finding. But such discoveries are hard to come by, and in many areas might actually never happen again—nothing guarantees that the fundamental discoveries in a given area have not already been made. ... This naturally leads to a publishing environment that pretty much mandates that findings are framed in the most favorable and exciting way, with important caveats and limitations hidden between the lines or missing completely. The author is too young to have directly experienced those times, but has read quite a few papers in top journals from the 1970s and earlier, and has been repeatedly struck by the difference between the open discussion one can find in many of those old articles and the currently dominant practices.

But that same problem is not limited to science itself, it seems to be now prevalent at all steps in the chain of transmission of findings, from the primary literature, through PR departments and press releases, and finally, in the hands of the science journalists and writers who report directly to the lay audience, and who operate under similar pressures to produce eye-catching headlines that can grab the fleeting attention of readers with ever decreasing ability to concentrate on complex and subtle issues. This leads to compound overhyping of results, of which The Deeper Genome is representative, and to truly surreal distortion of the science, such as what one finds in Nessa Carey’s Junk DNA.

The field of functional genomics is especially vulnerable to these trends, as it exists in the hard-to-navigate context of very rapid technological changes, a potential for the generation of truly revolutionary medical technologies, and an often difficult interaction with evolutionary biology, a controversial for a significant portion of society topic. It is not a simple subject to understand and communicate given all these complexities while in the same time the potential and incentives to mislead and misinterpret are great, and the consequences of doing so dire. Failure to properly communicate genomic science can lead to a failure to support and develop the medical breakthroughs it promises to deliver, or what might be even worse, to implement them in such a way that some of the dystopian futures imagined by sci-fi authors become reality. In addition, lending support to anti-evolutionary forces in society by distorting the science in a way that makes it appear to undermine evolutionary theory has profound consequences that given the fundamental importance of evolution for the proper understanding of humanity’s place in nature go far beyond making life even more difficult for teachers and educators of even the general destruction of science education. Writing on these issues should exercise the needed care and make sure that facts and their best interpretations are accurately reported. Instead, books such as The Deeper Genome and Junk DNA are prime examples of the negative trends outlined above, and are guaranteed to only generate even deeper confusion.
It's not easy to explain these things to a general audience, especially an audience that has been inundated with false information and false ideas. I'm going to give it a try but it's taking a lot more effort than I imagined.


1. Georgi Marinov is an author on the original ENCODE paper that claimed 80% of our genome is functional (ENCODE Project Consortium, 2012) and the paper where the ENCODE leaders retreated from that claim (Kellis et al., 2014).

ENCODE Project Consortium (2012) An integrated encyclopedia of DNA elements in the human genome. Nature, 48957-74. [doi: 10.1038/nature11247]

Kellis, M., Wold, B., Snyder, M.P., Bernstein, B.E., Kundaje, A., Marinov, G.K., Ward, L.D., Birney, E., Crawford, G.E., and Dekker, J. (2014) Defining functional DNA elements in the human genome. Proc. Natl. Acad. Sci. (USA) 111:6131-6138. [doi: 10.1073/pnas.1318948111]

206 comments :

1 – 200 of 206   Newer›   Newest»
NickM said...

"1. Georgi Marinov is an author on the original ENCODE paper that claimed 80% of our gnome is functional (ENCODE Project Consortium, 2012) and the paper where the ENCODE leaders retreated from that claim (Kellis et al., 2014)."

Given that...that last quote sure looks like a veiled critique of the 80%-functional claim and the ENCODE leadership. But...why be veiled? That's part of the problem! The right review would say straight-up the same kinds of things about that claim that it said in critique of these popular science books. If the real problem is in the peer-reviewed scientific literature, you have to fix it there first, or there is no hope in the popular literature!

Unknown said...

Georgi

To follow on from Nick's point above. Did you get any chance to see a draft of the manuscript that made the 80% functional claim before submission? Or was it just presented to you as a fait accompli by the big guys?

Larry Moran said...

Georgi has replied to this several times. He was only a graduate student at the time. Even if he had been given a manuscript (he wasn't) he was in no position to make waves. I understand that.

I think he deserves some credit for the second paper even though he doesn't agree with everything in it.

Larry Moran said...

I agree with your main point thatGeorgi should have been more forthright.

However, the peer reviewed scientific literature is never going to be perfect—nor should it. But there's no excuse for science journalists who don't do their homework.

Unknown said...

I do too. And I realise that administratively this would probably be inevitable. That was my point about the fait accompli question. Still, not a good way to do Science IMO

Georgi Marinov said...

When you write an academic paper, what matters is the subject at hand, not the people involved. I have hard time seeing how I could have said that I have been working for ENCODE without it being out of place.

Georgi Marinov said...

I don't know how that works in other fields where massive collaborations are common. I imagine that in high energy physics, for example, where everyone knows how to and does use LaTeX as a standard practice, they might have version control and it might be known who changed what and when all the way through.

But that's not how it works in biology -- there is a word file that initially gets sent to everyone (but sometimes it doesn't get sent at all -- there are a couple papers I am an author on which I knew were coming but never made it my way and I only realized I am on them when I clicked my name in PubMed), then people reply with edits, and after some point it may not even get sent to the whole group anymore. But even if it did get sent to everyone until the last draft, that doesn't really matter because it is impossible to have everyone agree on what the content is after each and every round of edits -- there are hundreds of people, most of them with all sorts of pressing demands on their time, more urgent than looking at endless drafts, and there is no way a paper can be published if each and every one of them had to be tracked down to sign off on the text. So what happens is that initially some relatively large fraction of the people involved look at the text, then more and more of them gradually just stop looking at it even if it does get sent to them, and eventually it's just a few senior people who work on the text (plus the journal editors) and everyone else has forgotten about the whole thing. And it is those people who determine what the final text looks like.

Steve said...

This critique of the critique is at most a pedantic distraction.

Regardless if not ALL non-coding was considered junk over the past 40 years, it is clear that Moran and like-minded acedemics emphatically support the idea that MOST non-coding DNA is still junk and will ALWAYS be junk.

So the critique of jDNA still stands.

These weak, pedantic distractions are a curious strategic choice of rhetorical weapons to try to stem the flow of information that supports an INCREASINGLY functional non-coding DNA reality.

gnomon said...

See what Prof. Ohta says about junk DNA in response to my direct question to her. And if you don't know who she is, you really have no business in things about junk DNA.

Shi Huang


> 发件人: "Tomoko Ohta"
> 发送时间: 2015-08-06 13:14:42 (星期四)
> 收件人: huangshi@xxxxx.edu.cn
> 抄送:
> 主题: RE: RE: Re: Re: EMBO Workshop
>
> Dear Dr. Huang,
>
> Thank you for the message.
> In the following, I put my response. I am in a hurry, and my sentence is not good.
>
>
> 1. what is your view on the junk DNA idea? If most variations are nearly neutral or slightly deleterious in humans, one cannot say that they are junks with no effects what so ever.
> Junk may not be truly junk. Recent knowledge on chromatin structure
> and function suggest that chromatin is very important for gene
> regulation. See for example, Nature Reviews Molecular Cell Biology
> 14:211-224 (2013)

Chris B said...

gnomon,

In your opinion, what percentage of the genome should be considered functional based on chromatin structure. and why?

Chris B said...

How does the level of junk DNA affect evolutionary theory or ID/creationist theory?

Mikkel Rumraket Rasmussen said...

@Gnomon

Onion test.

Larry Moran said...

The level of junk doesn't affect evolutionary theory at all.

It does affect Intelligent Design Creationism because the leading proponents of that movement have declared that their views predict the absence of junk DNA. Therefore, the presence of large amounts of junk DNA in our genome refutes Intelligent Design Creationism according to its own proponents.

Anonymous said...

Steve,

If we can recognize a large amount of the DNA as inactivated transposons, close to 50%, which is most likely true? That 80% of the DNA is functional, or that at least 50% is dead elements (junk)?

If we check the definition used by ENCODE for functional, and it comes back as a different definition to those that matter to biologists, most importantly those who have determined that if we find something dead it's junk, yet, dead elements would come back as "functional" by ENCODE's definition, should we dismiss the idea that dead elements are junk, or should we look carefully, and think better about what we're talking about?

If, say, introns were "necessary" because each and every alternative splicing variant was needed, would it mean that the whole thing, the whole intron, no matter its length, is necessary for this supposed function?

I disagree with Larry about the amount of junk, but I cannot disagree that it's a lot. My disagreement comes more from difficulties in clearly discriminating, for example, selfish elements from junk. Lots of pieces seem to be in-between (though I really not know how much of it). Dead elements can be transferred and duplicated by live elements, like the famous Alu elements. I don't know if I should call them selfish or junk, or active junk, or something else. In any event, both, selfish and junk DNA are not there to help us. They are just there.

Anyway, think before claiming stuff. Getting all triumphant because a lot of scientists inflate their results when they find some supposedly functional stuff proves nothing but that the system is becoming less and less academically serious. Soon your favourite creationist charlatans will be able to get away publishing abject crap, and we'll be left with a debate about how large the amount of junk ""science" is out there in academic journals. I think it's already above 80%. Maybe it's not junk, right? If people download them, even if they trash them after wards, it means they have to be functional articles.

Larry Moran said...

Steve says,

These weak, pedantic distractions are a curious strategic choice of rhetorical weapons to try to stem the flow of information that supports an INCREASINGLY functional non-coding DNA reality.

There are legitimate scientists who are opposed to junk DNA. The issue is still controversial although the consensus among knowledgeable experts is heavily in favor of lots of junk DNA.

As in any dispute, it's important to analyze all the facts and all the theories. That means you have to do your research and be aware of what your opponents are saying. It's not good enough to just promote the ideas and presumed facts that favor your side, you also have to acknowledge and refute the arguments on the other side.

Neither Nessa Carey nor John Parrington did that. That's why they are being criticized. It's not "pedantic," it's the essence of intelligent discourse.

But you probably don't know what intelligent discourse looks like, do you?

Anonymous said...

What? Is Ohta some kind of omniscient god? If not, why shouldn't we look at the data, rather than at somebody's opinion?

Jmac said...

"1. Georgi Marinov is an author on the original ENCODE paper that claimed 80% of our genome is functional (ENCODE Project Consortium, 2012) and the paper where the ENCODE leaders retreated from that claim (Kellis et al., 2014).


So...let me put it in simple language so that everyone, not really familiar with how real science works understands.

First, Georgi Marinov agreed with his science colleagues on ENCODE project that about 80% of human genome is functional; for reasons such as funding that is so hard come by (as he sited it himself).

Then, when Darwinists like Larry raised a hell that human genome can't be 80% functional, because they had predicted that it would be mainly junk, Georgi gradually changed his mind and begins criticizing everyone who supports the ENCODE 80% idea of human genome being functional.

But here is the best part: In order to look good, Georgi also begins to criticize the ENCODE papers directly, but in order not to look like an idiot (who criticizes himself) and a hypocrite, he decided to use veiled criticism, without actually mentioning the he was one of the authors of the papers that he is now criticizing.

I think most common people can be relieved that millions of dollars of hard-earned tax payers money was not wasted on some other studies where scientists now would be arguing and disagreeing whether 10 mg or 80 mg dose of a life saving drug should be administrated...

BTW: I have not read the two books that Georgi reviewed and so harshly criticized but can you imagine if both authors wrote their books or were inspired to write them, based on the articles that Georgi authored because they trusted they were accurate because Georgi was one of the authors?

Georgi Marinov said...

Seriously?

Jmac said...

Larry,

"The level of junk doesn't affect evolutionary theory at all.

Maybe, but it definitely affects evolutionary predictions made by evolutionary proponents like yourself. You continue to preach that most human genome will eventually turn out to be junk.

It does affect Intelligent Design Creationism because the leading proponents of that movement have declared that their views predict the absence of junk DNA. Therefore, the presence of large amounts of junk DNA in our genome refutes Intelligent Design Creationism according to its own proponents.

It only affects those proponents of the ID that insisted that the whole human genome would have to be functional. The vast majority of ID proponents are fine with some non-functional DNA due to various issues.

The only good thing about this whole issue is that people like you still insist that the majority of human genome is junk even though more and more experts gradually change their mind due to the emerging evidence that most of human genome is turning out to be functional.

Jmac said...

Just imagine that I'm one of the book authors who got inspired to write a book on junk DNA based on the papers you have authored.

Now, try to convince me why you would say that 80% of human junk was functional and now it was not.

Try to put yourself in my shoes... so to speak.

Georgi Marinov said...

Well, the first thing to do if you were going to write books based on papers is to read the papers. Not the NYT articles about the papers.

Neither the authors of these books nor you seem to have done that. You haven't read my review either. Which would be an absolute prerequisite for having the right to post what you wrote above

Chris B said...

Eric,

"The only good thing about this whole issue is that people like you still insist that the majority of human genome is junk even though more and more experts gradually change their mind due to the emerging evidence that most of human genome is turning out to be functional."

That's just false.

Chris B said...

Eric,

"Now, try to convince me why you would say that 80% of human junk was functional and now it was not. "

If you read the original ENCODE papers and the published responses to them, you would already have your answer.

Larry Moran said...

Maybe, but it definitely affects evolutionary predictions made by evolutionary proponents like yourself. You continue to preach that most human genome will eventually turn out to be junk.

I'm not making a prediction. I'm just saying that the sum of all the available evidence strongly supports the view that most of our genome is junk.

You are very confused about this evidence. You seem to think that we know nothing at all about genomes and the only reason we call it junk DNA is because of our ignorance.

If that were true, then my position would, indeed, be a "prediction." I would be predicting that all that mysterious dark matter will eventually turn out to be junk.

It's not mysterious dark matter. We have all kinds of evidence that it's junk. That's a conclusion based on facts and evidence, not a prediction.

Try and keep up.

judmarc said...

I believe you've done this before, but it would be good to see again a table with percentages in the human genome of known functional DNA, known junk (with types of junk), and the percentage that's still "up for discussion."

Alex SL said...

Ye gods, you have your name on several papers that you were not given a chance to read before publication? I am glad that it doesn't work like that in my part of biology...

Mikkel Rumraket Rasmussen said...

I love this exchange, it is instructive. Notice how we get a direct window to how an IDiot like Eric rationalizes the world and evidence around him into a preconceived conclusion.
Georgi tells us himself what happened, but that's not good enough for Eric because it doesn't fit into his creationist preconceptions about the atheist-Darwinist conspiracy.

He has this conspiracy theory in his mind with "Darwinists" controlling (and often faking) the science, and once in a while a naive researcher will overstep the boundaries and get put back in line.

They really do believe there's this sort of unofficial secret cabal of science-controlling atheist Darwinist boogeymen in control of everything. Amazing.

Diogenes said...

Eric says: "It [genome mostly junk] only affects those proponents of the ID that insisted that the whole human genome would have to be functional."

Which happens to be ALL of the most prominent ID proponents, and everyone who opined on the subject of DNA at the Discovery Institute-- except possibly Michael Behe.

But everyone else at the Discovery Institute, such as the Reverend Jonathan Wells, Cornelius Hunter, Casey Luskin, etc., are toast and their reputation is trash, or trash-ier than it would beotherwise.

Buh-bye Disco Tute.

Diogenes said...

Gnomon, nice novel you wrote about Georgi. You have a talent for fiction. Perhaps your talents would be better spent making a YouTube video on how a conspiracy of the Illuminati designed the Denver airport?

Diogenes said...

Gnomon, nice novel you wrote about Georgi. You have a talent for fiction. Perhaps your talents would be better spent making a YouTube video on how a conspiracy of the Illuminati designed the Denver airport?

Diogenes said...

Eric says: "It [genome mostly junk] only affects those proponents of the ID that insisted that the whole human genome would have to be functional."

Which happens to be ALL of the most prominent ID proponents, and everyone who opined on the subject of DNA at the Discovery Institute-- except possibly Michael Behe.

But everyone else at the Discovery Institute, such as the Reverend Jonathan Wells, Cornelius Hunter, Casey Luskin, etc., are toast and their reputation is trash, or trash-ier than it would beotherwise.

Buh-bye Disco Tute.

Fair Witness said...

It appears that Venter found a lot of junk, and he tossed it out. Good riddance !
http://news.yahoo.com/stripped-down-synthetic-organism-sheds-light-nature-life-204848392--finance.html

Diogenes said...

Larry, you should post this:

Modern Science explained briefly

John Harshman said...

Very nice, except that it overestimates the connection between the actual research and the university press release, which usually goes straight to "A causes B, proving Darwin wrong and revolutionizing everything we know about the universe".

BC said...

Professor Moran:

The level of junk doesn't affect evolutionary theory at all.

Professor Moran:

One of the things that Graur said was that if ENCODE is right then evolution is wrong. Now this may be a bit of an exaggeration, but not by much. If the ENCODE leaders are right that 85% of our genome has a biological function then we really do have to rethink the C-value paradox and genetic load. We also have to re-think a lot of biochemistry.

http://sandwalk.blogspot.co.uk/2013/07/what-did-dan-graur-say-in-chicago.html

txpiper said...

"It appears that Venter found a lot of junk, and he tossed it out. Good riddance !"

473 genes is not a short genetic profile.

Georgi Marinov said...

Those experiments have very little to do with the debate about junk DNA. They are extremely important on their own, but I don't see why anyone would bring them up in this context.

Georgi Marinov said...

They really do believe there's this sort of unofficial secret cabal of science-controlling atheist Darwinist boogeymen in control of everything. Amazing.

It's much easier to believe that -- at least if it was true, the world would be an orderly place that someone is in control of -- than to face the reality that it is one giant entropic mess that nobody can sort out and just drifts aimlessly along in spacetime.

Which is also a reason why religion has been so successful -- it gives the same reassurance (among other things).

txpiper said...

Right. Maybe doing experiments that involve stripping out all the supposed junk would be more telling.

Larry Moran said...

@BC

Yes, if 85% of our genome is functional and the DNA sequence is important then this is a problem for genetic load. That conflicts with evolutionary theory.

But we already know that the sequence of most of our genome is NOT conserved. The only possibility for function is one of the bulk DNA hypotheses that don't conflict with the genetic load argument.

I should have made that clear. If the function of most of our genome relies on the actual sequence of DNA then there IS a conflict with evolutionary theory.

Diogenes said...

Agreed, John. The cartoon needs to be revised.

The Other Jim said...

Isn't she simply saying that due to NNT, "proving" that some specified element is junk is not possible due to the greater effects of drift (due to Ne) as you approach "neutral"? (Assuming true junk-DNA is precisely neutral in terms of selection co-efficients)

Jmac said...

Let me get this straight Georgi;

Do you or don't you now claim that you initially supported the ENCODE's the 80% human genome functionality?

From Larry's posts all one could possibly infer is that you had initially supported the claim and then retracted your support for one reason or another. Either way is fine, I just would like to hear the story from your point of view.

Don't misunderstand me. I'm not judging you one way or another. There is politics everywhere.

My youngest son plays academy soccer U11 in Mississauga, Ontario, and there is politics there too. I have to sift through what everyone says and does because a lot of people selfishly look for their own interests and don't give a damn about anybody or anything else.

You could very well be one of those people who fell victim of the ENCODE ridden politics.

I just would like to get the facts from your point of view.
That's all.

Georgi Marinov said...

You still don't get it.

There never was a moment when the 460+ authors of that paper were asked "Do you agree with the claim that 80% of the genome is functional?". The vast majority of them only saw it on September 5th 2012 when the papers appeared online.

BTW, the paper itself does not even claim that -- it clearly defines what is meant by "biochemical function" and then says 80% of the genome was assigned some sort of "biochemical function".

And that was a very small part of the paper (a few sentences out of one of the longest papers Nature has ever published). What was said outside of the paper is where the claim that 80% of the genome is functional in the traditional sense of the term really originated. How much control do you think those 460+ people have over what editorials Nature puts together to advertise the ENCODE papers or what Gina Kolata writes in the New York Times?

Another thing you need to understand is that how much of the genome is "functional" was never really a major concern for ENCODE. I've been in countless conference phone calls, meetings, etc., it has never come up except for when the 2014 PNAS paper was being written (as a response to the reactions to the 2012 paper). People were doing functional genomics, and at a time when functional genomics was being transformed by the introduction of high-throughput sequencing. So the topics of discussion were experimental methods, computational analysis, etc. etc. And all that activity has moved science forward significantly (it does help when you have hundred of people working together every day rather than everyone hiding in their corner and their own little thing). There have been hundreds of papers produced by the consortium, and, as I said above, only one of them was explicitly about function, but that was a response to the reactions to a previous one, in which there were only a couple sentences about it.

But the creationists, with their minimal understanding of what science is and how it's done, think that ENCODE's sole purpose was to estimate how much of the genome is functional, and because it came up with an "inconvenient" answer, now it's backtracking to preserve the Darwinian orthodoxy. It's absolutely absurd.

William Spearshake said...

Larry, feel free to delete this, but I thought you should know that Denyse has banned Mapou from UD. First Joe, then Virgil (not Joe) and now Virgil.

But I guess it is too much to hope for that Barry is turning over a new leaf.

Larry Moran said...

Georgi,

The purpose of ENCODE was to define all the functional elements in the human genone. That's why you were looking for transcribed regions, transcription factor binding sites, DNase I sensitive sites, methylation sites etc etc.

I understand why the team spent most of their time worrying about the technology and the computational problems and that's perfectly normal in the middle of a difficult project. But maybe some of you forgot why you were doing all that work ... that's not normal.

The ENCODE leaders didn't forget why they were given so much money. It wasn't just to map all spurious transcription factor binding sites and junk RNA. They had to deliver on their promise to find out what all that genomic DNA is doing.

So they delivered by saying that 80% of the genome has biochemical activity knowing full well that the public would equate this with biological function. We know for a fact that many of the prominent leaders actually believed they had identified function in most of the genome. They certainly made no attempt to correct a "false" impression ... because they didn't think it was false.

They made similar claims back in 2007 when the results of the pilot project were published. The clear implication is that they were detecting function.

But the proof of the pudding is the (almost) complete absence of any critical analysis of their own results. There's almost no discussion of possible artifacts or alternative explanations in any of their papers. One gets the distinct impression that nobody working on the project even realized that they may be mostly mapping irrelevant, nonfunctional, sites. Most of them still won't admit this possibility, as you well know.

Understanding how much of the genome is functional was THE major concern of the project even though some of workers may have spent more time looking at trees instead of the forest.

Georgi Marinov said...

There is some misunderstanding here.

The purpose of ENCODE was to define all the functional elements in the human genone. That's why you were looking for transcribed regions, transcription factor binding sites, DNase I sensitive sites, methylation sites etc etc.

Yes, of course, nobody has forgotten that. What I was talking about is that there never was any discussion as in "We're going to say 80% of the genome is functional, does everybody agree?". And what people said individually cannot be taken to be the view of everyone who has been part of the project.

Larry Moran said...

And what people said individually cannot be taken to be the view of everyone who has been part of the project.

As a general rule, if your name is on the paper then it means you stand by all of the conclusions that are stated in the paper. That would certainly be expected if there were only two or three names on the paper.

As a general rule, if the paper contains results that you have produced then you are responsible for how those results are interpreted in the paper. You should stand by the conclusions and not try to cop out by claiming that the other two authors overruled you when the paper was being written.

Things get more complicated when dealing with hundreds of authors but it's fair to insist that all of the major lab PI's must defend the statements in the summary paper. If they don't then they should have withdrawn permission to use their data.

I suspect that all the ENCODE leaders agree on this point. They went along with the publicity campaign against junk DNA because they believed they had discovered function in most of the genome.

Some, but not all, backtracked 18 months later but the Kellis et al. paper looks more like an excuse than a mea culpa.

The truth about ENCODE

Georgi is not obliged to defend the original 2012 paper in spite of the fact that he is an author. However, it would have been nice to let readers know why he wasn't defending the paper.

Georgi Marinov said...

I actually discuss the problematic statement in the review. These are the key quotes:

These data enabled us to assign biochemical functions
for 80 % of the genome, in particular outside of
the well-studied protein-coding regions
...
Operationally, we define a functional element as
a discrete genome segment that encodes a defined
product (for example, protein or non-coding RNA)
or displays a reproducible biochemical signature
(for example, protein binding, or a specific chromatin
structure)
...
The vast majority (80.4 %) of the human genome participates
in at least one biochemical RNA—and/or
chromatin-associated event in at least one cell type


There is technically nothing wrong with these statements. Which I discussed. It was a huge mistake to write it that way, but the paper itself is not incorrect given that it provided a definition of "biochemical function".

The problem arose when everyone (and yes, that includes some people within ENCODE) forgot about how "biochemical function" was defined and started talking about just "function", and yes, that happened before the paper was even published.

Larry Moran said...

Georgi,

We've been over this before ...

What did the ENCODE Consortium say in 2012?

"Technically there's nothing wrong with these statements" may be literally true but it's still very misleading. Look at the whole paper, as I do in the post.

The authors could easily have said something like this ...

We recognize that some of the 'functional elements' we identify may not have a true biological function but we choose to call them 'functional' anyway because, as we say in the introduction, 'The Encyclopedia of DNA Elements (ENCODE) project aims to delineate all functional elements encoded in the human genome.'

The broader the definition of 'function' the better because it makes our results look revolutionary by overthrowing the idea of junk DNA.

But please, dear reader, don't be fooled by our rhetoric. We really don't mean it.

Georgi Marinov said...

"Technically there's nothing wrong with these statements" may be literally true but it's still very misleading.

Which is why I said it was a huge mistake do that.

Jmac said...

Georgi,

Thank you for your clarification. I will try to get back to it when I get a chance.

I'm asked this question a lot, so I figured I will ask you that as well; According to your current knowledge as well a the available research as of now, what is your prediction regarding what the human genome will turn out to be regarding "junk DNA"?

Georgi Marinov said...

Function isn't a binary characteristic, there is a continuum between vitally important stuff and pure junk. Still, the latter is the majority, around 85-90% of the genome in the case of mammals.

Georgi Marinov said...

Another review:

http://www.journals.uchicago.edu/doi/full/10.1086/685348

Larry Moran said...

A review of John Parrington's book by Elof Axel Carlson. Here's the most interesting paragraph of the review.

The first concern is the significance of the new epigenetics. What was called “junk DNA” in the early years of DNA sequencing is now seen by many biologists in a positive way. They argue genes can be switched on or off or modulated in intensity by mechanisms different from the operon model. The linearity of the genome implied by recombination mapping is not a simple “tape” to be played. Instead, the DNA often loops, linking distant religions together or it links regions of DNA in one chromosome with another nonhomologous chromosome. For Parrington, this supports a threedimensional rather than linear model of the active genome. The author assesses the arguments pro and con in a very objective way which is why I find this book so helpful.

I disagree. Parrington does NOT assess the arguments "pro" and "con" in a very objective way. In fact, he ignores most of the "con" arguments altogether.

Jmac said...

So, what you are trying to say is that the controversy between scientists who support your idea about junk DNA and the ones who think that the large of majority of genomes in mammals are functional or will turn out to be functional boils down to the interpretation of what biological function means? Or some groups of people insist on their interpretation of function?

I remember that Larry used to refer to the claim that "if you can remove large chunks of the genome and the organism is fine, this means that the chunks of genome that have been removed have no function and therefore they are junk".--more or less claim....

Do you support this view?

Larry Moran said...

Elof Axel Carlson also reviewed Nessa Carey's book in the previous issue of The Quarterly Review of Biology.

Here's an example of the quality of the review ...

The term junk DNA arose to describe most of the 98% of DNA that was not making protein. In the past 25 years, this inert or heterochromatic or nonprotein making DNA has collectively been called junk DNA. Nessa Carey provides us with the first popular overview of this field in 20 chapters. Her doctorate was in virology and she has an excellent grasp of the history of junk DNA and its implications for both basic and applied science.

Bullshit. Her grasp of the history is not excellent, it's abysmal. Reviewers should not comment on subjects they know nothing about.

Georgi Marinov said...

You are once again approaching the issue with a simplistic concept of "function" in mind.

Such experiments are unrealistic because one can never be certain that there is no phenotypic effect in the wild over prolonged periods of exposure to a wide variety of environmental conditions.

But in general, the most meaningful definition of "function" is the selected effect one, i.e. from the perspective of the organism segments of DNA are functional if they were historically selected for in the process of evolution and are maintained by purifying selection today. Note, however, that under this definition function is not a static characteristic -- say, for example, that you are a terrestrial mammal leading a lifestyle highly dependent on having a good sense of smell, and accordingly, you have a few hundred olfactory receptors. But then you move to an aquatic lifestyle and become a whale, and those receptors are no longer particularly useful, so there is no longer purifying selection acting on them, and they get pseudogenized, i.e. junk.

Also note that it is not possible to say that for example, exactly 310,045,614 bp of the genome are functional and the rest is not, because some positions are under very strong purifying selection, while others are barely above the drift threshold, and even if that was not the case, we can calculate "functionality" at varying levels of detail. For example, do we count all coding sequences or we exclude 4-fold degenerate synononymous positions, which can be anything? Even trickier is how we deal with UTRs -- there are functional elements within those, but they are more or less sparsely situated. Do we count enhancers as a whole or only the TFBSs within them? And do we count an 8-bp TFBSs as 8bp or we discount the 2 or 3 almost completely degenerate positions in the PWM? Etc., etc. One quickly realizes it is a futile exercise to try to derive accurate numbers for how much of the genome is "functional" because it does not even make sense to ask that question precisely. Not that anyone truly cares -- the primary reason people are having these discussions is creationists' obsession with junk DNA. The relevant scientific question is whether most of the human genome is "junk" and primarily shaped by neutral evolution or not. The answer is a resounding "yes, most of it is", has been so for 50 years, and has not changed at all in that time (and, BTW, would still be the same even if we were all created 6000 years ago by God and evolution was a lie, because the original arguments for why that is are in fact independent of the reality of evolution). Whether the best number for the fraction of the genome that is "functional" would be 9,10 or 11%, or even 5, 10 or 15%, is not really that important.

Identifying the individual functional elements are and figuring out what they are doing is a different story.

Georgi Marinov said...

I saw that one too, I didn't realize it was the same person...

P.S. Is it some sort of a taboo to publish a negative review of a book? I think the only ones I've seen are scientist attacking creationists and theologians attacking atheist scientists.

judmarc said...

the original arguments for why that is are in fact independent of the reality of evolution

Hard to say junk is defined as that which is not under selection (i.e., mutates at the neutral rate or something close to it), so you have mutation, selection and drift operating, then say all this works even if evolution is a lie. Trace the drift and selection back far enough and you get to common ancestry....

But I suppose it is more productive for me if I ask you what you meant, and then I will have the opportunity to learn something.

Joe Felsenstein said...

In most book reviews in science, the writer goes easy on the book's author. The author went to a lot of trouble to write the book, and if the facts in it are generally correct, one doesn't want to trash it even if it is partly unclear and even if the audience for which it is written is somewhat mysterious. One just makes some gentle remarks about how you did not understand the argument in Chapter 8, and at the end of the review about it not being clear who is the audience for this book.

Books that are pushing a dubious thesis, while ignoring evidence to the contrary and misleading unsophisticated readers are much more likely to draw a negative review.

Faizal Ali said...

I recently read a review of that Andrew Wakefield anti-vaccination movie that raised controversy when Robert DeNiro tried to get it shown at a film festival. The reviewer said he had no scientific background to judge the merits of the film`s claims and simply reviewed it on based its qualities as a movie. I thought that was a very ethical approach. (He didn't like it).

Larry Moran said...

Georgi asks,

Is it some sort of a taboo to publish a negative review of a book?

We live in a postmodernist age where everyone's opinion is assumed to be valid and true. American society is also very reluctant to tolerate anything that seems to be a personal attack if it's directed against an "average" person.

Those two things combine to make it difficult to write a scathing review of a book. It means you can get away with publishing all kinds of nonsense. It means that anyone can give a TED talk and get a loud applause for saying really stupid things. It means that any stage performance gets a standing ovation.

Larry Moran said...

lutesuite writes,

I thought that was a very ethical approach.

It's the same kind of "ethical approach" used to award book prizes. You can win a Pulitzer or another major prize for a science book as long as it's well-written and fun to read. It doesn't matter if the science is crap.

I don't think we should tolerate that kind of "ethics." As I've said repeatedly, the three most important criteria in popular science journalism (and teaching) are" 1. accuracy, 2. accuracy, and 3. accuracy.

When you review a book, your primary obligation is to review it's accuracy. Nothing else matters. If the facts are wrong then the book is crap no matter how well it is written.

The same applies to a movie or a documentary.

Jonathan Badger said...

Larry, "Postmodernism" hasn't been much of a thing since the 1990s. You can attack current trends in the humanities, but rather than saying "everyone's opinion is valid", the trend today is to talk about "privilege" -- people certainly can criticize other people's opinion, but only if you aren't seen as "punching down".

Faizal Ali said...

@ Larry:

Don't get me wrong. It would have been best if someone knowledgeable of the relevant science had been sent to review the film. But since I doubt the reviewer has the say on which subjects he is assigned, the best he could do would be to acknowledge his ignorance.

A science periodical, OTOH, has a higher responsibility to assign knowledgeable reviewers IMHO. And for a reviewer to feign knowledge he does not possess is unpardonable.

Faizal Ali said...

Here is the review, BTW:

http://www.hollywoodreporter.com/review/vaxxed-cover-up-catastrophe-film-879960

Jmac said...

You are once again approaching the issue with a simplistic concept of "function" in mind.

What's wrong with that? Would making things look more complicated change the issue that many scientists, not only Larry, used knockout experiments to prove their point of view that the majority of mammal genome must be junk?

Such experiments are unrealistic because one can never be certain that there is no phenotypic effect in the wild over prolonged periods of exposure to a wide variety of environmental conditions.

I agree with your statement. I can assure you, there is more to the story than that.

But in general, the most meaningful definition of "function" is the selected effect one, i.e. from the perspective of the organism segments of DNA are functional if they were historically selected for in the process of evolution and are maintained by purifying selection today.

You are making an assumption here, which may or may not be viewed as bias...

Note, however, that under this definition function is not a static characteristic -- say, for example, that you are a terrestrial mammal leading a lifestyle highly dependent on having a good sense of smell, and accordingly, you have a few hundred olfactory receptors. But then you move to an aquatic lifestyle and become a whale, and those receptors are no longer particularly useful, so there is no longer purifying selection acting on them, and they get pseudogenized, i.e. junk.

Really?

Jmac said...

2.

Note, however, that under this definition function is not a static characteristic -- say, for example, that you are a terrestrial mammal leading a lifestyle highly dependent on having a good sense of smell, and accordingly, you have a few hundred olfactory receptors. But then you move to an aquatic lifestyle and become a whale, and those receptors are no longer particularly useful, so there is no longer purifying selection acting on them, and they get pseudogenized, i.e. junk.

Really?

Here we go again... This is one of the typical evolutionary approaches to junk DNA Georgi. I'm not going to comment on that yet, but I have to ask you this question first; since the organism that evolved and changed the environment of life why natural selection keep the chunks of DNA that are no longer needed? You don't have to answer this question! It is pointless and we both know it.

Is it just in case they need to be revived one day if the mammal returns to live on land again?

Also note that it is not possible to say that for example, exactly 310,045,614 bp of the genome are functional and the rest is not, because some positions are under very strong purifying selection, while others are barely above the drift threshold, and even if that was not the case, we can calculate "functionality" at varying levels of detail. For example, do we count all coding sequences or we exclude 4-fold degenerate synononymous positions, which can be anything? Even trickier is how we deal with UTRs -- there are functional elements within those, but they are more or less sparsely situated. Do we count enhancers as a whole or only the TFBSs within them? And do we count an 8-bp TFBSs as 8bp or we discount the 2 or 3 almost completely degenerate positions in the PWM? Etc., etc. One quickly realizes it is a futile exercise to try to derive accurate numbers for how much of the genome is "functional" because it does not even make sense to ask that question precisely.

Would you agree that some, even influential evolutionists, over-used this argument to promote atheism, evolution and their own egos? I have many names...

Not that anyone truly cares -- the primary reason people are having these discussions is creationists' obsession with junk DNA.

I don't! Do I have explain why?

The relevant scientific question is whether most of the human genome is "junk" and primarily shaped by neutral evolution or not. The answer is a resounding "yes, most of it is", has been so for 50 years, and has not changed at all in that time

The only thing I can tell you Georgi is that if the neutral evolution is true, it is not the Darwinian evolution, and you MUST HAVE moved on with people like Larry who realized that defending Darwinism is no longer sustainable. But you did! Will you do the same when "junk DNA" bebecomes a player? I will track you down when this happens... believe me!

(and, BTW, would still be the same even if we were all created 6000 years ago by God and evolution was a lie, because the original arguments for why that is are in fact independent of the reality of evolution). Whether the best number for the fraction of the genome that is "functional" would be 9,10 or 11%, or even 5, 10 or 15%, is not really that important.

Aren't you contradicting your own statement in the second piece? You have assumed the evolutionary influence, and now you are saying that it would not be important!
Can you see the problem here Georgi?

Identifying the individual functional elements are and figuring out what they are doing is a different story.
Well... that may very well be true but is prevention considered to be a "biochemical function" in your view?

Jmac said...

I forgot this piece by Dan Graur:

"If ENCODE is right, evolution is wrong"!

Yeah... ENCODE had nothing to do with junk DNA but evolutionists think otherwise.

Georgi Marinov said...

"Postmodernism" hasn't been much of a thing since the 1990s.

Because it has become the commonly accepted position so there is no need to discuss it. At least that's my impression as an outsider.

Georgi Marinov said...

What's wrong with that? Would making things look more complicated change the issue that many scientists, not only Larry, used knockout experiments to prove their point of view that the majority of mammal genome must be junk?

Such experiments an demonstrate that most of the genome is junk. What they cannot show is how much exactly. Which is what I was talking about

Georgi Marinov said...

Here we go again... This is one of the typical evolutionary approaches to junk DNA Georgi. I'm not going to comment on that yet, but I have to ask you this question first; since the organism that evolved and changed the environment of life why natural selection keep the chunks of DNA that are no longer needed? You don't have to answer this question! It is pointless and we both know it.

Is it just in case they need to be revived one day if the mammal returns to live on land again?


It is not at all pointless to reply.

The answer is that they are not kept, they do eventually go away. The question is how long it takes for that to happen in the absence of selection against their presence.

They are most definitely not "kept for the future", evolution has no such foresight.

Georgi Marinov said...

Would you agree that some, even influential evolutionists, over-used this argument to promote atheism, evolution and their own egos? I have many names...

No, I don't agree.

The only thing I can tell you Georgi is that if the neutral evolution is true, it is not the Darwinian evolution, and you MUST HAVE moved on with people like Larry who realized that defending Darwinism is no longer sustainable. But you did! Will you do the same when "junk DNA" bebecomes a player? I will track you down when this happens... believe me!

Evolutionary theory in the 21st century is not Darwinian. What is your point?

I forgot this piece by Dan Graur:

"If ENCODE is right, evolution is wrong"!


That's a hyperbole.

First, ENCODE has not claimed selected-effect functionality for the whole genome at the sequence level, which is what that statement implies.

Second, if there was selected-effect functionality at the sequence level for the whole genome, that would mean that our understanding of population genetics and evolution is not right, because we have gotten the parameters horribly wrong. That would not mean that evolution did not happen though, only that the theory we have is wrong (or at the very least, the parameters of it). Here, we come back to how one has to distinguish between evolution as the fact of common descent with modifications and evolution as the theory that explains how that happened.

Larry Moran said...

ENCODE supporters aren't bothered by the fact that most of the genome isn't conserved. They think that huge amounts of regulatory sequences and regulatory RNAs have evolved quite recently. The lack of conservation (selected-effect function) isn't a problem, it's a feature. It's how they explain the wonderful complexity of humans in spite of the fact we have the same number of protein-coding genes as a rat.

If that doesn't work, they can always fall back on one of the bulk DNA hypotheses that ascribe function that's not dependent on sequence.

Georgi Marinov said...

They think that huge amounts of regulatory sequences and regulatory RNAs have evolved quite recently. The lack of conservation (selected-effect function) isn't a problem, it's a feature.

But something that has evolved recently and has been selected for in order to create the complexity of humans would still fall under the selective-effect function category, wouldn't it?

Then of course we start once again talking about onions, salamanders and dinoflagellates :)

Larry Moran said...

But something that has evolved recently and has been selected for in order to create the complexity of humans would still fall under the selective-effect function category, wouldn't it?

Yes, of course it would. It would be a selected function with no evidence to support the claim of selection. In other words, indistinguishable from DNA that is NOT selected.

That was my point.

Arno Wouters said...

Georgi, Larry and judmarc, in connection with 'junk DNA' it is important to distinguish Georgi's verbal definition of function as selected effect from Larry's notion of function as being important in the life of the organism (I call the latter 'function as biological role'). It seems to me that it is the absence of function as biological role that makes something into junk.

Actually, I feel that although Georgi says that he defines 'junk' as absence of selected function, he, in fact uses the term in the sense of absence of biological role. This is most clear from the following remark:

The relevant scientific question is whether most of the human genome is "junk" and primarily shaped by neutral evolution or not. The answer is a resounding "yes, most of it is", has been so for 50 years, and has not changed at all in that time (and, BTW, would still be the same even if we were all created 6000 years ago by God and evolution was a lie, because the original arguments for why that is are in fact independent of the reality of evolution)

As judmarc suggests, if 'junk' is defined as having no function as selected effect, the answer to the question of how much DNA is junk would not be independent of the reality of evolution. After all, if evolution was a lie, no segment of DNA would have a function as selected effect. So, if someone thinks that the arguments for junk DNA are independent of the reality of evolution they probably think of function as biological role, rather than as selected effect.

I would agree that the arguments for junk are independent of the reality of evolution, but in my eyes this indicates that 'junk' means 'no biological role' rather than 'not under selection'.

I would think that Georgi's verbal notion of function according to which "segments of DNA are functional if they were historically selected for in the process of evolution and are maintained by purifying selection today" is useless in the context of the junk dna debate. If a gene that codes for an essential protein does not vary in the population it would be very strange to say that that segment is junk. Yet, it would be junk under Georgi's definition because it is not currently under selection (there are no heritable differences in fitness related to this segment).

More importantly, if absence of purifying selection would be sufficient to conclude that a segment isn't junk, the amount of junk dna would be irrelevant to the question whether the genome is "primarily shaped by neutral evolution or not" for it could easily occur that a segment that has been shaped by natural selection is not currently under selection.

On the other hand, if junk is defined as lack of function as biological role, the existence of large amounts of junk dna strongly supports the thesis that the genome is primarily shaped by neutral evolution, for it would be very strange if segments that are not important in the live of the organisms were shaped by selection.

For that reason I think that in the context of the junk DNA debate junk is better defined in line with Larry's view as lack of function as biological role than as lack of a selected function.

hoary puccoon said...

If "a gene that codes for an essential protein does not vary in the population" couldn't that be because every organism that carries a mutated version dies early in development?

Georgi Marinov said...

Yes, of course it would. It would be a selected function with no evidence to support the claim of selection. In other words, indistinguishable from DNA that is NOT selected.

There would be evidence -- it would be constrained within the population.

Arno Wouters said...

hoary puccoon, I am not a biologist and I am eager to hear what relevant experts have to say about your question. However, I understand from the literature that it is extremely improbable that all possible mutations of a protein coding gene are lethal. Most protein coding genes contain introns that can mutate without affecting the protein. Synonymous substitutions are typical neutral or nearly neutral.

Even in cases where every possible mutation would be lethal, it is not clear to me how that would count as the segment being maintained by purifying selection as long as such mutations do not occur.

Moreover, the possibility that every mutation of the gene for an essential protein is lethal is not that relevant to my argument that in the context of the junk dna debate function is better defined as not having a biological role than as absence of selection. The point of my reference to an invariant gene for an essential protein is that no one would bother about checking whether or not selection occurs if it is clear that a gene codes for an essential protein in order to conclude that it is functional.

Jonathan Badger said...

It's been twenty years since the Sokal Social Text affair. Postmodernism lost. Even the few remaing holdouts like Bruno Latour got freaked out when creationists and global warming deniers started using pomo arguments.

SRM said...

It's been twenty years since the Sokal Social Text affair.

Is that the definitive dividing line you imagine? In any case, such "movements" tend to leave significant residue. Morph as it will.

Joe Felsenstein said...

I went to a history-of-science . meeting in Berlin in 2010 called Writing Genomics The title of the meeting implied that different views were just different "texts". There were a moderate number of people there who used postmodern language. I was mentally preparing myself for a fight, in which I would argue that the very confidence with which they sat in the room and did not expect the imminent collapse of the roof meant that science did actually make progress.

However despite some unconscious self-parody along the lines of "we must wonder about the validity of this sentence", I found little conflict. A nonpostmodernist historian I talked to about it said that he thought that the postmodernists had lost and at some level knew it. Of course they have left some effects, such as an increasing willingness of all to recognize that their views are not "objective" but are strongly affected by the societal milieux.

I also got the impression that postmodernism is stronger in the U.S. than in Europe, which surprises me since it originated in Europe, I think.

Joe Felsenstein said...

Correction: The meeting was in 2008.

Corneel said...

@Arno Wouters
Hoary puccoon meant to say that the very fact that there is little genetic variation at a certain gene is in fact strong evidence of ongoing purifying selection. You should pick another example :-)

Corneel said...

@Arno Wouters
For essential proteins both definitions of function coincide, because the absence of function automatically results in purifying selection. The problem is of course deciding what is a genuine "function", when the effect is less dramatic. If you are interested, this issue has been extensively discussed by Ford Doolittle.
http://sandwalk.blogspot.nl/2013/03/ford-doolittles-critique-of-encode.html

hoary puccoon said...

Corneel @ 8:50:00 AM--

Precisely.

Arno Wouters said...

Thanks for your replies, Corneel!

Perhaps I should clarify the background of my remarks. Let me start by emphasizing that I am not advocating the view that the ENCODE results give us reason to abandon the idea that most of our genome is junk. I agree with Larry and Georgi (and Doolittle) that that view is bonkers.

Doolittle (in the paper referred to by Corneel) and Graur et al. (in their "On the immortality of television set", Genome Biology and Evolution 5(3):578–590, 2013) reproach ENCODE for not having used what they think is the proper notion of function namely 'function as selected effect'.

What I wanted to point out in the above comments is that it is not needed to define function in terms of selection to see that most of our genome is junk. If 'functional' is defined (in line with Larry's proposal) as 'having an important role in the survival and reproduction of the organism' and junk as 'not functional is this sense' there is still no reason to think that ENCODE has shown that most of our genome is junk.

Georgi points out that the arguments for junk DNA are independent of selection. This gives me the impression that his verbal definition of function as selected effect is at odd which the way he uses the term 'function' (namely as biological role). I am very curious to hear what he thinks about this.

To my mind, the definition of function in terms of selection is unfortunate.

Doolittle boldly presents his definition of function as selected effect as a defense of the traditional understanding of function. He provides no argument for that view and he is wrong. Since the sixteenth century 'function' has mainly been used to refer to either the activity or the role of parts of organisms in the maintenance of an organism. The notion of function as selected effect was invented by philosophers in the 1980s for the purpose of a naturalistic philosophy of mind and language, and, to my mind, doesn't help in understanding biology. I argue this in my "Four Notions of Biological Function" (2003).

However, the main reason why I am worried about the tendency of some evolutionary biologists to define function in terms of selection is that this definition all to easily leads to a confusion of claims about organisms and claims about populations or lineages and hence to a lazy form of biology in which the performance of a biological role is seen as sufficient evidence for selection.

The claim that a certain segment of DNA codes for an essential protein is a claim about the role of that gene in the life of the organism. For the truth of this claim it doesn't matter whether or not this gene varies in the population and how much these variants differ in fitness.

The claim that a certain segment of DNA is under purifying selection is a claim about the population. For the truth of this claim it matters a lot what other variants there are in the population and how much their fitnesses differ. If there are no variants or if their differences are neutral or nearly neutral there is no purifying selection.

The automatic inference from performing an essential role to currently under purifying selection is, in my opinion, therefore, a case of lazy biology.

Georgi Marinov said...

Georgi points out that the arguments for junk DNA are independent of selection.

I pointed out that some of the arguments for junk DNA are independent of evolution being true. That's a different thing.

Arno Wouters said...

Georgi, that's a different thing for sure. I was a bit inaccurate in my representation of your view in my comment of Monday, April 04, 2016 2:29:00 PM. Sorry.

But what do you think of my original argument (Sunday, April 03, 2016 2:25:00 PM):

As judmarc suggests, if 'junk' is defined as having no function as selected effect, the answer to the question of how much DNA is junk would not be independent of the reality of evolution. After all, if evolution was a lie, no segment of DNA would have a function as selected effect. So, if someone thinks that the arguments for junk DNA are independent of the reality of evolution they probably think of function as biological role, rather than as selected effect.

And can you agree with my view that in order to avoid confusion of claims about the organism and claims about selection and lazy conclusions about selection on the basis of claims about organisms it is better not to define function in terms of selection? If not, why do you think it important to define function in terms of selection?

Georgi Marinov said...

Not really.

Evolution as a fact = common descent, old Earth, etc.

Evolution as theory explains how it happens.

It is perfectly possible for a deity to have specially created everything 6000 years ago, i.e. there is no common descent, but from that moment on natural selection operates and certain regions of the genome are strongly contrained by it while others are not. And it would have to be that way because there is no other option given the known mutation rates, the size of the genome and the characteristics of eukaryote genes. Natural selection is not really a theory, it's more of a truism, i.e. a direct and unavoidable consequence of the relationship between genotype and phenotype and of differential reproduction.

Georgi Marinov said...

P.S. There is in fact quite a bit of independence between the fact of evolution and evolutionary theory.

It is possible that common descent is true as we know it, but that it was all guided by the deity, i.e. the theory is false because mutations and gamete segregation are not in fact random as it assumes.

It is also possible that the theory is true but common descent is completely false, i.e. the world was created specially 6000 years ago. But its development after that is described by the theory very well because the deity does not intervene at all and the assumptions of random mutations and gamete segregation hold.

There is no reason to think any of these hypothetical possibilities matches reality, but they are important thought experiments nevertheless.

judmarc said...

Thanks Georgi, I understand what you meant now.

judmarc said...

@Arno Wouters - Not Georgi, but I'll try.

Function and selection are not really different criteria, with the exception of any function one might find that was unaffected by mutation (e.g., fish with "antifreeze" made of junk DNA, "bulk DNA" arguments). Other than such unusual examples, where the genetic code is pertinent to function it will tend to be preserved by selection where currently beneficial, or change rapidly under the influence of selection where a new mutation is beneficial; where it is not pertinent to function, it will tend to vary per the "molecular clock," i.e., not be strongly selected for or against.

Arno Wouters said...

Thanks Georgi for the explanation! I now see how one can think that in absence of evolution DNA segments can have selected functions. I also see that genetic load and related arguments are not independent of the existence of selection: if all mutations were neutral or nearly neutral these argument would not work.

However, I still don't see that these arguments require a definition of function in terms of selection. Do you think they do?

Arno Wouters said...

judmarc, thanks for your attempt. The idea that function as biological role and selection are not really different criteria is in my view the kind of misunderstanding that is fueled by thinking of function in terms of selection. It is true that segments of dna can only be selected for if they have a biological role. This means that conservation is strong evidence both for biological role and past selection. However the reverse is not true, a gene can have a biological role, without being under selection, namely if there is no existing variation (I have no idea how often this occurs - I would be thankful if knowledgeable people would enlighten me on this) or if all existing variation is neutral or nearly neutral (my evolution books tell me that this is quite common). For that reason, I think that in the context of the junk DNA debate function is better defined in terms of biological role than in terms of selection (and, as I said before, it is also more in line with the biological tradition).

Jmac said...

"Function isn't a binary characteristic, there is a continuum between vitally important stuff and pure junk. Still, the latter is the majority, around 85-90% of the genome in the case of mammals.

Really? Well, you probably haven't read some of the papers that summarize many experiments with mice. What the experiments have found is that mice's so called junk DNA works as night vision. This is just one of many examples of the so-called junk DNA having a function.

Is prevention a function in your opinion Georgi?

Georgi Marinov said...

It's safe to say that it is not me who is misinformed here

Aceofspades said...

Larry, one thing which I think is worth discussing is the way Venter's team recently reorganised the genome of their new organism (syn 3.0). They effectively showed that the organisation of genes doesn't need to be as haphazard as we find it to be in living systems.

See here

Christopher Voigt, Professor of Biological Engineering, MIT wrote:

Evolution has given us this mess of genetics that works, but is not easy to understand or manipulate. So, just like organizing a house, they took the different genes that encode different functions in the cell and started to sort them out, putting all the genes associated with a particular pathway in the same portion of the genome.

If human designers can create an ordered, structured alternative to how life is found in nature, that would speak to the complexity of biology simply being an artefact of how it was shaped by evolution.

This is a fairly strong argument against design. A living system that was designed should look ordered. This has once again demonstrated that the synteny is somewhat irrelevant (there are some exceptions of course) and so there is no reason why a designed system should not look designed.

judmarc said...

a gene can have a biological role, without being under selection, namely if there is no existing variation (I have no idea how often this occurs - I would be thankful if knowledgeable people would enlighten me on this) or if all existing variation is neutral or nearly neutral

Let's think more about these two situations.

- No variation: Since no DNA is exempt from mutation (shielded from cosmic rays and background radiation, never exposed to environmental mutagens), if we find no variation, then to me that can mean just one thing: Very strong purifying selection must be at work. Can you think of another circumstance where there will be no variation over time?

- All variation is neutral or nearly so: And how do we learn that all variation is neutral or nearly so, since we are a long way from knowing what specific biological function(s) each particular stretch of DNA in the genome has or hasn't got? Why, because it shows no or nearly no signs of selection. Can you think of a situation in which a stretch of DNA has an important function in the biology of the organism but is not under any sort of selection (aside from the rare situations where sequence doesn't matter)?

judmarc said...

What the experiments have found is that mice's so called junk DNA works as night vision.

Thus more rats survived and caused the Black Plague in Europe, killing tens of millions, proving the existence of a merciful God who designed all this.

The Other Jim said...

Queuing the "cannot know the mind of the designer" argument in 3....2.....1

Chris B said...

Eric,

"Well, you probably haven't read some of the papers that summarize many experiments with mice. What the experiments have found is that mice's so called junk DNA works as night vision. This is just one of many examples of the so-called junk DNA having a function."

And what would 'intelligent design theory' predict about the levels of junk DNA in diurnal rodents in light of these data?

Aceofspades said...

I'm attending a debate on ID in 2 weeks. I'll raise this point in Q&A afterwards and I'll let you know how accurate your prediction was.

Arno Wouters said...

judmarc (April 05, 2016 6:32:00 AM) wrote:

Since no DNA is exempt from mutation (shielded from cosmic rays and background radiation, never exposed to environmental mutagens), if we find no variation, then to me that can mean just one thing: Very strong purifying selection must be at work. Can you think of another circumstance where there will be no variation over time?

My evolution books tell me that there are several other possible causes of lack of genetic variation. The main ones are genetic bottlenecks and founder effects.

I have heard about several case of very low variability in almost every gene.

The most extreme example I know of is the conifer *Wollemia*, which has virtually no genetic variation. Other examples are: the Serbian population of Golden jackals, cheetahs, European bisons.

It would be absurd to say that the genes of these organisms do not have functions. Yet, this is precisely what one should say if function is defined in terms of selected effects ("segments of DNA are functional if they were historically selected for in the process of evolution and are maintained by purifying selection today").

By the way: as far as I know mutation due to cosmic rays, background radiation and environmental mutagens is not important in evolution.

Can you think of a situation in which a stretch of DNA has an important function in the biology of the organism but is not under any sort of selection (aside from the rare situations where sequence doesn't matter)?

As I said, as far as I know, a lot of the existing variation within functional genes is neutral or nearly neutral because it occurs in parts of the gene that are irrelevant to its function or because it concerns synonymous substitutions.

Georgi Marinov said...

My evolution books tell me that there are several other possible causes of lack of genetic variation. The main ones are genetic bottlenecks and founder effects.

Yes, but those usually result in absence of variation genome-wide. While the relevant thing here is absence of variation in some regions of the genome relative to others

judmarc said...

Hello Arno -

We appear to be talking about different things.

First, variation: Yes, of course you're correct that replication errors are, generally speaking, almost undoubtedly responsible for more mutations than any of the factors I named. Same point, though - *over time*, given that replication is not perfect, what can explain low incidence of variation at a given locus? I'm speaking of classic stuff like the Hox gene cluster. What can explain its low variation over time other than selection?

Your examples are of particular incidences of low variability *within a given population at a specific time*, and you've provided some reasons for that (another, in the case of Wollemia, is that like aspens it clones itself). But this is not at all the same topic as determining whether a particular genomic locus, over eons and species, changes more or less rapidly. If it hardly varies, then the reason is purifying selection, unless you want to theorize that some species' cells have learned the secret of perfect replication. And where there is purifying selection there is function.

Second: I asked whether you could think of a situation where a genomic locus has an important function but is not under any sort of selection, and you responded with examples where change would be irrelevant to function. And here is where, just as expected, we would see variation occur at a neutral or nearly neutral rate. So: Not under selection, irrelevant to function. Agreed?

judmarc said...

And what would 'intelligent design theory' predict about the levels of junk DNA in diurnal rodents in light of these data?

Eric's example has nothing to do with junk DNA. He doesn't realize that, but there is no reason we must be similarly uninformed.

The diurnal mammal example illustrates a DNA function that actually does depend on sequence (there is a particular arrangement of it that has the light-channeling function), *but it is not a genetic function*. DNA is simply serving as a light-focusing material; it is not coding for protein, regulating development, etc. So this is a little like the "fish antifreeze" example, but in contrast to the fish, sequence matters here.

If one wanted to draw a lesson from this, I suppose it would be that evolution doesn't do planning and design, it doesn't spec the best materials from a list, it works with the materials it's given.

Arno Wouters said...

judmarc, Georgi and Corneel, thank you for helping me to clarify my thoughts!

As judmarc observes, I wasn't talking about lack of variation in a specific DNA segment in the course of time (I would call that 'conservation' rather than 'lack of variation'). I was talking about the lack of variation in a specific DNA segment of individuals that co-exist ('compete') in a population (I shall abbreviate that as 'synchronic variation' to avoid further confusion).

I don't dispute that past selection is the only reasonable explanation of conservation and I agree that variations that are irrelevant to function are not under selection.

So, yes if you were talking only about conservation, we were talking about different things.

I brought up the lack of synchronic variation because, as far as I know, selection is defined as the existence of heritable fitness differences between co-existing ('competing') variants in a population. This means that if there is no synchronic variation for a specific DNA segment or if that synchronic variation is neutral or nearly neutral that DNA segment is (by definition) not under selection.

So if 'function' is defined in the way Georgi and Doolittle do ("segments of DNA are functional if they were historically selected for in the process of evolution and are maintained by purifying selection today") and junk as lack of function in that sense this would mean that:

- the DNA of Wollemia is 100% junk (in other words: Wollemia doesn't have genes)
- the dna-segment of European bisons that codes for the HBA1 polypeptide has no function, whereas the same segment of the American bison has.
- human dna-segments that code for essential proteins and that currently do not synchronically vary or whose current synchronic variation does not result in more than nearly neutral differences in fitness don't have a function.

If not absurd, this is at least clearly at odds with the way in which biologists use the term 'function'.

So the way in which the term 'function' is used indicates that 'function' (and, hence, 'junk') are not used in the sense of function as selected effect, but rather in the sense of having a biological role.

Furthermore, if (as Georgi and Doolittle maintain) being functional would imply currently under purifying selection, in order to establish that a certain dna-segment is functional, it would neither be sufficient to show that that segement codes for an essential protein nor that it has been conserved (that is selected in the past) but it should also be shown that that segment varies synchronically in a non-neutral way in the current population. The fact that nobody thinks that such a check is needed is another indication that the relevant notion of function is function as biological role rather than function as selected effect.

Do we agree?

John Harshman said...

- the DNA of Wollemia is 100% junk (in other words: Wollemia doesn't have genes)

Did you just divide the genome into protein-coding DNA and junk? Don't let Larry find out.

Arno Wouters said...

Did you just divide the genome into protein-coding DNA and junk? Don't let Larry find out.

I just didn't, I never did so, and would not dare to do so!

What I am saying is that under Georgi's definition of junk as not being under selection even the protein coding parts of Wollemia's DNA would be junk. Which is one of the reasons why I prefer Larry's definition of junk as not important in the life of the organism over Georgi's.

hoary puccoon said...

Arno--

Are you testing "synchronic variation" on organisms at or directly after their conception? If so, how? If not, how do you know they aren't being rigorously de-selected long before birth? If you test organisms that are already born, you've got them only after they've passed the most rigorous part of the selection process. If you test only mature individuals, then you've ignored even more of the selection process. It doesn't mean there wasn't variation in the original population. It means you've only tested the individuals who have already passed the test.

John Harshman said...

I just didn't, I never did so, and would not dare to do so!

You did, but I'm sure you did it by accident. When you said "in other words", you divided the genome into genes and junk.

I know what you meant to say. I think the argument is silly, for many reasons.

Arno Wouters said...

Hoary, you'd better address those questions to Georgi. He is the one who defines function in terms of selection. One of the benefits of the notion of function as biological role is that there is no need to worry about such tests.

judmarc said...

What I am saying is that under Georgi's definition of junk as not being under selection even the protein coding parts of Wollemia's DNA would be junk. Which is one of the reasons why I prefer Larry's definition of junk as not important in the life of the organism over Georgi's.

What this does is confusingly conflate what you are thinking of as selection (via competition among individuals in a population) and genetic loci that are "under selection," which is the sense that both Georgi and I are using. To quote Georgi: "[T]he relevant thing here is absence of variation in some regions of the genome relative to others."

This is the sense in which both Georgi and I are using "selection," and as we have already discussed, it is in this sense that "under selection" is equivalent to "having function," while "varying at the neutral rate" is, with very few exceptions, equivalent to "junk."

Unknown said...

To add to judmarcs comment above - it makes sense to compare genomes of related species and check whether a particular region is more conserved than others. During a short time period there simply may not be any novel mutations in that region (which becomes less likely if populations are large). But if you are comparing related species, you are basically increasing your sample size by taking into account a large number of generations and what you detect are regions that have been conserved by selection whenever novel variants appeared.

Chris B said...

Arno,

Why would Wollemia's protein coding DNA be considered junk?

Why would the European bison's HBA1 peptide be junk?

Chris B said...

judmark,

"If one wanted to draw a lesson from this, I suppose it would be that evolution doesn't do planning and design, it doesn't spec the best materials from a list, it works with the materials it's given."

I just wanted Eric to give me a rational explanation for these things under ID/creationist theory. If ID/creationists want to be taken seriously as presenting a scientific alternative to evolutionary theory, they are going to have to actually present their theoretical ideas at some point with regard to specific problems, not vague generalities.

Corneel said...

Arno Wouters asked
Do we agree?

No, we do not. You are trying to find a situation where the definition of function as "selected effect" runs into trouble. But even if there is currently no genetic variation in a population, genes will still be under purifying selection, because you can't stop the introduction of new deleterious mutations.

I don't see any obvious disadvantage of using the "selected effect" definition of function. In fact, it is equally trivial to find an example where function as biological role would give some problems. Consider for example this thread where Larry discussed transposable elements. What, if any, function do they have? Is "generating novel variation" a valid function since it is beneficial to the population, not the organism (which violates your definition)? And how would you test that function without resorting to selected effect arguments?

Arno Wouters said...
This comment has been removed by the author.
Arno Wouters said...

judmarc and Corneel, are you saying that I erroneously equate 'currently under selection' with 'currently being selected for'? If so, how would you define 'currently under selection'?

Judmarc, you mention "the absence of variation in some regions of the genome relative to others". However, relative absence is at most an indication of past selection. How would it make sense as a definition of "currently under selection"?

Furthermore, because all individual Wollemia trees have virtually identical genomes, it is probably not the case that their functional segments show less variation that their non-functional ones. So, on this definition of 'under selection' the whole genome of Wollemia would be junk.

Corneel, are you suggesting that 'a DNA segment is currently under selection' should be defined as 'some mutations in that segement would be deleterious' or 'deleterious mutations in that segment would be selected against?" It seems to me that in both cases it applies to a lot of DNA segments that are (in my opinion, and probably in yours too) junk, because it is nearly always possible that segments of junk DNA mutate in a way that has deleterious effects. So I guess that under this definition the genome of Wollemia doesn't contain any junk.

Georgi Marinov said...

all individual Wollemia trees have virtually identical genomes

They don't.

Pretty much each cell in a multicellular organism has a unique genome. Keep that in mind.

Georgi Marinov said...

Let me try to clear the confusion here.

The most simple-minded definition of "function" is "stuff that matters". Stuff that matters is by definition under selection, otherwise it wouldn't matter. The great majority of the time that selection is purifying. In a few odd cases it might be promoting sequence diversity, but those are rare and do not change the overall picture much.

There is literally almost nothing to argue over here, it is mostly tautological truisms, so I have hard time understanding the purpose of this back-and-forth.

Georgi Marinov said...

it is nearly always possible that segments of junk DNA mutate in a way that has deleterious effects.

Yes, it is. Let's say a mutation creates a TFBS in a otherwise neutrally evolving segment that disrupts the regulation of a nearby gene.

There is still a significant difference between that segment and a real enhancer -- most mutations in it would have no effect, while a special one creates a negative effect. In contrast, many more mutations in the real enhancer will have negative effects and will be selected against.

Arno Wouters said...

Pretty much each cell in a multicellular organism has a unique genome. Keep that in mind.

Thanks Georgi that clarifies much! So, you are suggesting that in order to determine what parts of a certain genome are functional and what not we should compare the genomes of thousands (how many are enough?) of cells of the same individual and see what parts vary?

I like it that in this way the term 'functional' is tied to the level of the individual (which is IMHO where it belongs), but how would this be indicative of being currently under selection?

Corneel said...

@Arno Wouters
In addition to the excellent remarks of Georgi:

You said ..
...it is nearly always possible that segments of junk DNA mutate in a way that has deleterious effects
Yes, but the definition that was introduced earlier in this thread specified that a locus had to have experienced some historical selection for its current function in order to qualify as a functional sequence.

...are you saying that I erroneously equate 'currently under selection' with 'currently being selected for'?
The issue here is the meaning of "current" in evolutionary time. No genome remains free of mutations for a long time on the timescales we are considering here.

Corneel said...

I like it that in this way the term 'functional' is tied to the level of the individual (which is IMHO where it belongs)
So driving elements and genomic parasites are not functional sequences by your preferred definition?

Alex SL said...

Is it possible that there is some confusion going on here between the underlying definition for "functional" and a criterion that is used to infer that a region fulfils that definition? Kind of like the difference between "acidic" and "showing a low pH on my pH-meter"?

Unknown said...

In other words, the Disk Cleanup program has never been used in our hard drive genome...

Jmac said...

By PZ Myers on May 19, 2010

"The ENCODE project made a big splash a couple of years ago — it is a huge project to not only ask what the sequence of a strand of human DNA was, but to analyzed and annotate and try to figure out what it was doing. One of the very surprising results was that in the sections of DNA analyzed, almost all of the DNA was transcribed into RNA, which sent the creationists and the popular press into unwarranted flutters of excitement that maybe all that junk DNA wasn't junk at all, if enzymes were busy copying it into RNA. This was an erroneous assumption; as John Timmer pointed out, the genome is a noisy place, and coupled with the observations that the transcripts were not evolutionarily conserved, it suggested that these were non-functional transcripts".

Can anybody recognize what the main argument of PZ. Myers was? Not the actual data but the assumption from bias evolutionary point of view, that those transcripts would have to be non-functional.

All I can say Georgi Malinov is that for the rest of your career you have to assume that evolution is true first and then fit whatever data into your predisposed paradigm.

Jmac said...

Obviously, there is more assumption to be taken if you want to succeed if you are a new graduate like Malinov. You need to gradually disown yourself, as well your common sense and judgment. You can't survive in this crowd if you have a divided heart.

Starting with the origin of the universe with the mind-boggling evidence that it had to have begun fully organized, to the origins of life that nobody can replicate. To proving evolution is a fact in the lab; I have a few projects I can assign but so far, nobody wants to do it, to prove evolution to be a fact.
Malinov, would you like to make a name for yourself and do some real research instead of criticizing people who wrote crappy books thanks to you?
I don't have unlimited funds but I could possibly squeeze you in, since you mentioned that "junk DNA" is left-over of functions of mammals who moved into the aquatic environment.
Would you like to be involved in such an endeavor?

SRM said...

This is why Larry refers to people like you as IDiots. No one ever declares that a given transcript has to be non-functional.

The Myers quotation (which I think you must have transcribed badly) makes it clear what the issue is and Larry has also discussed the pervasive/spurious transcription issue here many times. But you don't read and you don't comprehend.

Chris B said...

No, Eric, those transcripts don't have to be non-functional, but when they are present on average <1 copy per cell and come from regions on non-conserved DNA, if you want to claim they are functional the burden of proof is on you.

judmarc said...

on this definition of 'under selection' the whole genome of Wollemia would be junk.

What this shows is that trying to determine what is under selection from an instantaneous snapshot of an unusual situation results in nonsense.

Here's Wollemia, an organism that does at least some (most?) of its reproduction by making clones of itself. It's at a dead end, very few left at only a single location, having been wiped out over most of its range. Not at all surprisingly, there is very little genetic variation among the trees in this population. So what overarching conclusions for the whole of biology should we draw about the definitions of genetic variation and function from this instantaneous snapshot of the last few isolated members of a species heading toward extinction?

None.

Evolution occurs over time, not instantly. In a population large enough and over enough time, variation will occur; if it is in a functional part of the genome, it will be selected for or against (usually the latter); if it is in a part of the genome that is junk, it will usually not be subject to selection at all (barring something occurring within the "junk" that affects something functional, which is unusual but does happen). Over thousands upon thousands of generations, three groupings of genetic variations over time will show up: (1) faster than neutral rate changes will occur in functional sequences where there are favorable mutations; (2) in "junk" the rate of mutation will be at the neutral "molecular clock" rate; and (3) in functional sequences where unfavorable mutations have occurred from time to time in some members of the population, the rate of mutation will be slower than the neutral rate due to purifying selection.

This is just population genetics math.

Anonymous said...

Eric,

The reason nobody cares about your "offer" is that you're an idiot. You do not understand what you're talking about. The worst is that you think you do understand something, but you're just self-deluded, and out of reach from reason.

judmarc said...

with the mind-boggling evidence that it had to have begun fully organized

Oh dear. Would you like to tell us what writings of an IDiot said this and convinced you, or alternatively which actual scientist you are misunderstanding?

judmarc said...

All I can say Georgi Malinov is that for the rest of your career you have to assume that evolution is true first and then fit whatever data into your predisposed paradigm.

I think whenever you see the word "evolution" or a derivative, you stop thinking. The reason I say this is because I don't believe even you would be stupid enough to miss the clear implications of what you bolded above unless you simply hadn't bothered to think about it.

Don't you understand that when PZ Myers says the transcripts weren't conserved, it means that other species don't produce these same transcripts, and thus if they had an important biological function these other species couldn't exist? But they do exist, which shows the transcripts serve no necessary function.

Ed said...

Eric wrote:
" To proving evolution is a fact in the lab; I have a few projects I can assign but so far, nobody wants to do it, to prove evolution to be a fact. "

Ah yes, that project, the one where you want to perform a mass murder on creationists, right? And you're surprised no one is taking it?

And your religious morals are supposed to be superior to atheist morals?

The Other Jim said...

...and who needs anti-virus software.

Petrushka said...

If cloning or parthenogenesis were automatically a risky method of reproduction, dandelions would be endangered. It is interesting to ponder why the designer made dandelion pollen.

Larry Moran said...

The PZ Myers post is on Panda's Thumb: Junk DNA is still junk.

He links to 2007 blog posts by John Timmer and by me: ENCODE finds the human genome to be an active place, What is a gene, post-ENCODE?.

Note that these posts refer to the preliminary ENCODE results on 1% of the human genome. They were published in 2007. All three of us make the same point; namely, that most of the transcripts are rare and their sequences are not conserved. Given those facts, the onus is on ENCODE to provide evidence that they are functional. The default assumption is that they are just noise since we know from decades of work on transcription that the process is error prone.

There were many more critiques of ENCODE back in 2007 and the criticisms were perfectly valid. The ENCODE Consortium leaders choose to ignore all those criticisms when they published the complete study in 2012.

That's not how real scientists are supposed to behave.

The main point in PZ's article was to highlight a new (2010) paper by my colleagues Ben Blencowe and Tim Hughes here at the University of Toronto. They showed that one of the techniques used to make the claim for pervasive transcription was flawed. Their work indicated that less than half of the genome produces stable transcripts that could be functional.

See my post at Junk RNA or Imaginary RNA? and Carl Zimmer's post at How Many Sparks in the Genome?.

judmarc said...

If cloning or parthenogenesis were automatically a risky method of reproduction, dandelions would be endangered.

Getting back to the original topic, it sure does cut down on genetic variation between individuals (unless you're a microbe doing HGT), doesn't it? :-)

It may not be decisive in itself, but when combined with other "risk factors" (e.g., limited population numbers and geographic range, definitely not true of dandelions) it could represent a strike against ultimate survival.

judmarc said...

One other comment, going back to some things Arno mentioned early on: If Wollemia manages to survive its present straits, we'd look back at this a few thousand years from now as a bottleneck, and there would be signs in the plant's genome of that bottleneck, just as there are signs in the human genome of population bottlenecks Homo sapiens went through. Bottlenecks, founder effects, etc., are not unusual in the history of species. This is not at all surprising when thinking of the reproductive isolation necessary for the creation of new species. Such signs are detectable because they occur against the background determined by selection/function.

Arno Wouters said...

Corneel (Wednesday, April 06, 2016 4:58:00 AM):

You are trying to find a situation where the definition of function as "selected effect" runs into trouble.

Actually, no. I am trying to find situations in which the notion of function as biological role and function as selected effect lead to different verdicts.

judmarc (Wednesday, April 06, 2016 6:58:00 PM):

So what overarching conclusions for the whole of biology should we draw about the definitions of genetic variation and function from this instantaneous snapshot of the last few isolated members of a species heading toward extinction?

To avoid misunderstandings, I am not trying to draw overarching conclusions for the whole of biology from instantaneous snapshots. What I am trying to do is to draw conclusions about how biologists use the term 'function'.

Arno Wouters said...

Corneel (April 06, 2016 10:00:00 AM)
So driving elements and genomic parasites are not functional sequences by your preferred definition?

Sure, they aren't.

Arno Wouters said...

Georgi (Wednesday, April 06, 2016 9:01:00 AM):
There is still a significant difference between that segment and a real enhancer -- In contrast, many more mutations in the real enhancer will have negative effects and will be selected against.

An interesting solution that didn't occur to me!

How are you going to count the number of possible negative mutations?

SRM said...

My apologies to Eric then, regarding my comment that he must have badly transcribed the Myers quotation: it is exactly as it was written. Myers doesn't usually write in such a ham-handed fashion, I don't think.

Jmac said...

I really like reading sandwalkblog because in most cases it provides me with ALL the information I need to know about a particular subject. It sometimes amazes me by providing the reader with accurate information and facts.
The ENCODE issue is no different, surprisingly to some.
On Larry's blog "How does Nature deal with the ENCODE publicity hype that it created?" we find ALL the information we need about the issue. I love it , so let's begin.
I don't have to write much as Larry is pretty accurate at explaining as to why he and the team of devote Darwinists hate the ENCODE.
Let's see:
Larry begins his argument that the video by leading ENCODE scientists is a hype.
You be the judge:

https://www.youtube.com/watch?v=Y3V2thsJ1Wc


http://sandwalk.blogspot.ca/2014/05/how-does-nature-deal-with-encode.html

Jmac said...

To sum this up; What ENCODE is actually saying, they are not backing out of their initial statements back in 2012 that 80% of human genome has a biological function, as Larry confirmed himself here:

http://sandwalk.blogspot.ca/2014/05/how-does-nature-deal-with-encode.html

1. 80% plus of human genome has biological function but that can change to 100%
2. Why it can change to 100%? Because ENCODE, unlike Larry, Malinov and Graur, they only studied 147 cell types. When they study more cell types like Larry, Malinov and Graur did in their lab experiments, they sure will come up with the same number...

Jmac said...

But to really, truly understand the issue, one has to look at the two sides agendas; what do they really care about? What is their main goal?

A neutral/unbiased observer should have no problem at all to identify what each side CARES ABOUT THE MOST?

CAN YOU?

Chris B said...

What are the "two sides"?

The 80% functional vs. the ~10% functional genome? Or the ID/creationist vs. science in terms of evidence for junk DNA?

Arno Wouters said...

judmarc Wednesday, April 06, 2016 6:58:00 PM:

(2) in "junk" the rate of mutation will be at the neutral "molecular clock" rate; and (3) in functional sequences where unfavorable mutations have occurred from time to time in some members of the population, the rate of mutation will be slower than the neutral rate due to purifying selection.

You aren't saying that junk DNA attracts mutations, are you?

This is just population genetics math.

Is it? Or is it confusion of substitution rates with mutation rates?

Chris B said...

Arno,

"You aren't saying that junk DNA attracts mutations, are you?"

No. Mutations occur at roughly the same rate throughout the genome. In DNA that is not under sequence contraint, these random mutations can accumulate. In sequences subject to selection because they are functional, deleterious mutations will be selected against.

"Is it? Or is it confusion of substitution rates with mutation rates?"

Even if one were confusing the two, it would not change the fact that judmarc is talking about basic population genetics math.

Arno Wouters said...

I would appreciate it if someone would point me to a basic population genetic math. text that discusses the difference between 'being currently under selection' and 'currently being selected for or against'.

Arno Wouters said...

Chris B.
In sequences subject to selection because they are functional, deleterious mutations will be selected against.

If 'being functional' is defined in terms of selected effect 'subject to selection' is just another word for 'functional', so under that definition it can't be the case that a sequence is subject to selection because it is functional.

Actually, this is one of the reasons why I think it is important to distinguish function and selection. 'Function' is an important explanatory concept in biology (and has been so since the end of the 18th century). It is (among other things) used to explain why selection occurs. But as selection is oblivious to the past, if 'function' is defined as 'that for which it was selected in the past' function cannot explain selection. If, alternatively, function is defined as 'currently being selected for' function cannot explain selection either because in that case saying that something is functional is just another way of saying that it is currently selected for, not an explanation.

I explain this in my Biology's functional perspective.

Besides, as I said before, I believe that deleterious mutations will be selected against even if they occur in fragments that are not considered functional. Georgi (Wednesday, April 06, 2016 9:01:00 AM) seems to agree with me.

Mikkel Rumraket Rasmussen said...

@Eric

The ENCODE people all believe in evolution too. So tell me, what does each side "CARE ABOUT THE MOST" ?

Arno Wouters said...

Corneel and judmarc, in clarification of / addition to what I said in my comment of Thursday, April 07, 2016 1:36:00 PM:

I don't think the notion of function as selected effect is in trouble in the case of Wollemia etc. On the contrary I think it is clear that if function is defined in terms of "currently being selected for " (in addition to being selected in the past) no part of the DNA of Wollemia has a function (even stronger, I think that under this definition none of its parts has a function). The point is not that the notion of function as selected effect is in trouble in the case of Wollemia. The point is that if you cannot agree that none of the DNA of Wollemia has a function you are not using the term 'function' in a way that implies 'currently being selected for'.

I think we agree on this, because judmarc (Tuesday, April 05, 2016 9:03:00 PM) pointed out that Doolittle's and Georgi's definition doesn't talk about "currently being selected for" but about "currently being under purifying selection".

I am trying to understand how something can be currently under selection (purifying or other) if there is no current synchronic variation. So once again my question is can you point to an explanation of 'currently under purifying selection' that is generally accepted in population genetics and doesn't require synchronic variation?

Futuyma's Evolution (2013) p. 310/11, defines 'purifying selection' as directional selection in which the disadvantageous allele is lowered in frequency and may be entirely eliminated. This definition clearly requires there to be disadvantageous alleles. In the case of Wollemia there is no synchronic variation, so there are no advantageous or disadvantageous alleles.

Arno Wouters said...

Is it possible that there is some confusion going on here between the underlying definition for "functional" and a criterion that is used to infer that a region fulfils that definition? Kind of like the difference between "acidic" and "showing a low pH on my pH-meter"?

Yes, that's possible. Thank you.

I'll use this distinction to clarify myself:

I think that when talking about junk DNA (as well as in many other contexts in biology) the underlying definition of 'functional' (cp. 'acidic') is something like 'performing an important role in survival and reproduction'. 'Sequence conservation' is one of the criteria that indicate that a sequence is functional in this sense (cp. "showing a low pH on my pH-meter"). Coding for a product with a known biological role is another one.

Let me add (to avoid misunderstandings) that I think that ENCODE's notion of biochemical function is a valid notion of function (as valid as function as biological role and function as selected effect) but that biochemical function isn't the same as function as biological role. I think that while it can be the case that, given their definition of biochemical function, the ENCODE consortium was able to "to assign biochemical functions for 80 % of the genome" (I don't have the expertise to judge that claim), it is surely not the case that their results give us reason to doubt that most of our genome is junk (because, as I said, the relevant notion in the junk DNA debate is function as biological role).

I also agree with Larry and Georgi (and many others) that ENCODE's claims about function are very misleading.

I furthermore think that the papers of Graur et al. and Doolittle added to the confusion (rather than clarify it), because they introduced the irrelevant notion of function as selected effect into a discussion that was already confused.

In my opinion Germain, Ratti and Boem do a good job to unravel the confusion in their Junk of Functional DNA?. I know that Graur thinks otherwise.

Chris B said...

Arno,

"If 'being functional' is defined in terms of selected effect 'subject to selection' is just another word for 'functional', so under that definition it can't be the case that a sequence is subject to selection because it is functional."

Functional in this case would mean that a sequence has a biological function that affects the fitness of the organism, and is therefore subject to natural selection.

"Actually, this is one of the reasons why I think it is important to distinguish function and selection. 'Function' is an important explanatory concept in biology (and has been so since the end of the 18th century). It is (among other things) used to explain why selection occurs. But as selection is oblivious to the past, if 'function' is defined as 'that for which it was selected in the past' function cannot explain selection."

Of course selection is oblivious to the past. But past selection leaves its mark on sequences that have historically been under selection. Selection is also not an omnipresent entity or force that applies itself under certain circumstances. Selection happens whenever there is sequence variation and those sequences variants have differential fitness in that population. It is a consequence of these circumstances.

"Besides, as I said before, I believe that deleterious mutations will be selected against even if they occur in fragments that are not considered functional. Georgi (Wednesday, April 06, 2016 9:01:00 AM) seems to agree with me."

I agree with you here also. But in order for selection to happen, the sequence experiencing selection must be having some fitness effect on the organism, otherwise there is no differential survival and therefore no selection.

Chris B said...

Arno,
"I would appreciate it if someone would point me to a basic population genetic math. text that discusses the difference between 'being currently under selection' and 'currently being selected for or against'."

How would you explain the difference between these two phrases?

judmarc said...

Arno, you're leaving me behind. :-) What you're discussing appears to be more complex, and I'd just like to pause at this point and try to express my point of view (as a non-expert layperson who could be talking out of his hat) as simply as I can.

- As the discussion surrounding ENCODE has shown, it can be difficult to assess whether part of the genome is functional (that is, whether it performs an important role) simply by reference to things like whether it is transcribed.

- In order to help with the assessment of what is functional (in terms of whether it plays an important role) it can be very helpful to look at the historical context of a species' genome.

- Are portions of the genome quite well conserved over aeons and species? Then it is likely they are functional, i.e., they play an important role in the species' reproduction and/or survival. A classic example would be something like the Hox cluster. We would say this cluster is so essential to proper development and thus survival that it is under very heavy purifying selection.

- Are portions of the genome undergoing change considerably faster than the "background" rate, compared to other species that share a recent common ancestor? Then we would say it is likely this part of the genome is under positive selection. (An example might be genes involving speech in humans.)

- What about the portions of the genome that are changing at the neutral/background/molecular clock rate? We would say in terms of historical context that these portions of the genome are not under either positive or purifying selection. We may further conclude the reason we do not see this portion of the genome under selection is because it is neither important to reproduction and survival as it stood in ancestral species (thus no purifying selection), nor does it have a recent change that confers advantage in reproduction and survival (thus no positive selection). Because it has no role in reproduction and survival, i.e. no function, there is no historical context of selection.

- Of course things are never quite this neat and simple. There may be less than the typical amount of variation that should exist in accordance with the background/neutral/molecular clock mutation rate due to founder effects, population bottlenecks, reproduction via cloning, etc. But we see the signs these things have occurred in the organism's evolutionary history against the background of the rest of its genome that has either been under selection or has changed at the neutral rate.

- Regarding our old friend Wollemia, if we see today as it sits in its isolated geographical location with little genomic variation between individuals, must we throw up our hands and declare its entire genome suddenly has converted entirely to junk or entirely to function? I don't think so. When we talk about selection being equivalent to function, we are talking about the context that is given by examination of changes over evolutionary time scales. To gain that context for Wollemia, we would have to examine other species that share recent common ancestors with it. The fact that the species shows little variation among individuals at this instant in time shows only that it has come to either the point of extinction, or, if it survives, a population bottleneck (which would leave telltale signs in the genome for biologists examining Wollemia tens of thousands of years in the future).

Struggler said...

Why do creationists get so melodramatic , it just ups the insanity factor

Unknown said...

eric-
A test —
Here is a link to an experiment that shows Darwin was wrong-
http://www.nsf.gov/discoveries/disc_summ.jsp?cntn_id=131308

Have the researchers who published this been forced to recant? Has their funding dried up? I don’t know, but what you find out might be interesting.

Unknown said...

Aceofspades
What Venter’s team did is interesting and amazing, but I don’t think they accomplished what you seem to be claiming they did.

Ventner’s team did not create a structured alternative to life found in Nature. What they did was create a chimera that can reproduce under very carefully controlled conditions. It wouldn’t make it outside the lab. What they did might lead to a life form that would be a structured alternative to what we find in nature, but they have not done so yet.

One possible difficulty—
http://science.sciencemag.org/content/351/6280/aad6253.full

“Unexpectedly, it also contains 149 genes with unknown biological functions, suggesting the presence of undiscovered functions that are essential for life.”

That’s a significant part of the genome that is not understood. Until that is better understood claims of ‘good design’ or that an ordered genome that would make it outside the lab are premature.

christine janis said...

How has this NSF funded study disproved common descent?

SRM said...

Simple fact is that these religious culture warriors are steeped like a forgotten tea bag in an agenda. They therefore imagine a scenario where scientists doing science are likewise motivated by religious agendas.

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

I was hoping that someone would sum it up what the real, main point what this "controversy" is all about... I will try to do it the best I can:

Quick facts:

1. ENCODE came up with the number of 80% that there were able to assign to biochemical functions, either specific or nonspecific, to the parts of the human genome previously known or called as "junk DNA."

2. Quite a few of evolutionary biologists became very vocal about the findings of ENCODE, calling it and their authors names, basically: names.

3. Few evolutionary biologists even put their reputation on line in order to criticize the ENCODE and make them retract their claims: One of them is the host of this blog as well as:
PZ. Myers, Dan Graur, Nick Mitzke, T.R. Gregory, Jeff Schallit, Georgi Marinov (I apologize for misspelling your last name Georgi in my last post or ever. My tablet often changes the spelling without me noticing it.) and Larry Moran being the most vocal and confident about the ENCODE findings being wrong as well as Nature magazine blowing it up and creating the hype along the the medias.

Well, this may not be the best summary I have ever written but the parties involved in the discussion know what I'm talking about. At least, I hope so.

4. ENCODE did not retract their original claims as of today, (as Larry also pointed it out clearly in his blog linked by me above as of 2015).

ENCODe also continue to insist, that the 80% number could very well go to 100%; meaning, that further studies could prove that the human genome is 100% functional, i.e. there is not junk DNA at all...

So...? Big deal a neutral observer would possibly say. How is that a problem?

5. The real issue is that many evolutionary biologists have made predictions and still insist on those predictions (names above) that the human genome would have to have useless chunks of DNA just decaying, because the process of evolution just couldn't keep up with the removing of it fast enough, or something like that...

Now, the findings of ENCODE put that prediction in shambles...big time.

So, what's a scientific theory worth if it can't predict anything? Well...

However, the interesting part of it all is that, as far as I know, the pillars of the prediction on junk DNA; PZ. Myers, Dan Graur, Nick Mitzke, T.R. Gregory, Jeff Schallit and Larry Moran are still standing tall; they are so convinced that ENCODE is wrong, they put their reputation, life achievements and in many cases everything they have known for many, many years one line.

BTW: I'm surprised and curious at the same time why J. Coye has been avoiding the discussions we have been having...

Unknown said...

So, I've waited for a while hoping that somebody else might take this on, but it has to be me again, does it?

different views were just different "texts"

There is a somewhat technical way in which "text" is used here and it makes as much sense to put the word "just" in there as it does to talk about "just a theory". A similar confusion appears for "social construct". We can not communicate about anything that's not a social construct, because a social construct is the relation between a signifier and that which it signifies (for instance the word "tree" and an actual tree). If we want to talk about something, we first need to agree upon the meaning of words (hence we socially construct the means of communication). "Text" is simply a collection of social constructs. That's broader than simply language - a figure is a text even without the caption. Everything we can discuss is text.

I would argue that the very confidence with which they sat in the room and did not expect the imminent collapse of the roof meant that science did actually make progress.

I doubt you have the notion of progress postmodernist have critiqued. But really if you want to understand this point, you have to start by looking at the notion of progres modernists were promoting and then at the criticism leveled against that particular notion. A modernist account of science holds that science started with some questions and as time went on generated more and more accurate answers. The postmodern criticism of this view is that it fails to note that novel insights also produce novel questions. It's not as if we merely had a better understanding of molecular phylogenetics than - say - Newton had, it's that Newton didn't even have the basic notions required to understand molecular phylogenetics available to him. If this seems a lot like Simpsons criticism of describing horse evolution as a linear progression rather than taking into account the diversification of the clade it's because it is the same point. The "ladder" view of evolution was modernist ideology superimposed on biology.

If we are talking about postmodern approaches to history, there is one feature that makes them different from both pre-modern and modern views: The recognition of contingency. Modern and pre-modern views either thought of history as premeditated (fate, divine guidance, the deterministic laws of history in historical materialism) or as intentionally shaped by actors.

We live in a postmodernist age where everyone's opinion is assumed to be valid and true.

This also misses the point. Relativism holds that there is more than one consistent (and thus valid) view. That does not mean that all views are valid (and in fact a lot of views are not). Since we don't have any tools by which we can pick one consistent view over another we get relativism.
1+1=2 in many valid arithmetic systems.
1+1=0 in Zmod2 (which is consistent and actually surprisingly important)
But AFAIK there are no consisten arithmetic systems in which 1+1 is any other number. There are two correct answers (depending on which system you are using), but there are a lot of wrong ones.

Chris B said...

Eric, sigh, you really can't create the reality you want.

"I was hoping that someone would sum it up what the real, main point what this "controversy" is all about... I will try to do it the best I can:"

You failed.

"Quick facts:"

No, slow narrative assembly by ID/creationist proponents that you have slavishly regurgitated on cue:

"1. ENCODE came up with the number of 80% that there were able to assign to biochemical functions, either specific or nonspecific, to the parts of the human genome previously known or called as "junk DNA."

No. They said that ~80% of the human genome was transcribed into RNA,most of which was at <1 copy per cell and from regions of DNA under no detectable sequence constraint. Given what we have know for decades about the nature of DNA transcription, this should not have been terribly surprising.

"2. Quite a few of evolutionary biologists became very vocal about the findings of ENCODE, calling it and their authors names"

They became vocal for the reasons I explained above, which you've heard many times. Characterizing their published critiques as 'namecalling' is wrong. Since you are repeating this after being corrected, you are lying.

"3. Few evolutionary biologists even put their reputation on line in order to criticize the ENCODE and make them retract their claims: One of them is the host of this blog as well as:
PZ. Myers, Dan Graur, Nick Mitzke, T.R. Gregory, Jeff Schallit, Georgi Marinov"

Another lie, Eric. Many scientists called out ENCODE for their methods. However, some scientists, not familiar with ENCODE's publication, became aware only after its Kardashian style hyping in the media, and pouncing of ID/creationists on the bogus functionality claim.

"and Larry Moran being the most vocal and confident about the ENCODE findings being wrong as well as Nature magazine blowing it up and creating the hype along the the medias."

Wrong again. While Dr. Moran has been a major proponent in pointing out the obvious flaws and inconsistencies in the ENCODE 80% functionality paper, many other scientists have weighed in on this fiasco. You have yet to address any of their arguments, instead regurgitating ID/creationist pablum.

"Well, this may not be the best summary I have ever written but the parties involved in the discussion know what I'm talking about. At least, I hope so."

I would say this is an average summary of what you have to offer on this subject and the variation about the mean is really, really small.

"4. ENCODE did not retract their original claims as of today, (as Larry also pointed it out clearly in his blog linked by me above as of 2015)."

Another delusion, Eric. They backed off big time from their functionality claim.

"ENCODe also continue to insist, that the 80% number could very well go to 100%; meaning, that further studies could prove that the human genome is 100% functional, i.e. there is not junk DNA at all..."

Dr. Marinov can dispel you of such fantasies, Eric, but if it helps you sleep, keep dreaming.

"So...? Big deal a neutral observer would possibly say. How is that a problem?"

IDK, Eric, you tell me. Who has a problem here with junk DNA. ID/creationists, or science?

Chris B said...

Eric,

"So, what's a scientific theory worth if it can't predict anything? Well..."

Good point. Name one thing ID/creationist theory predicts, and how it is consistent with ID/creationist theory

Arno Wouters said...

Judmarc, I'm relieved how well you take my remarks! Thank you.

I can agree with most of the views expressed in your summary.

Regarding our old friend Wollemia, if we see today as it sits in its isolated geographical location with little genomic variation between individuals, must we throw up our hands and declare its entire genome suddenly has converted entirely to junk or entirely to function? I don't think so.

I like your humorous way of expressing this!

I completely agree that the lack of variation in Wollemia populations doesn't indicate that its genome has converted entirely to junk or entirely to function. Actually, that is one of the reasons why I think that the relevant notion of function is function as biological role.

However, I don't think that there is currently much selection going on in Wollemia, at least not if selection is defined as it is in most textbooks of evolutionary biology namely as the existence of heritable differences in fitness between variants within a population.

The fact that the species shows little variation among individuals at this instant in time shows only that it has come to either the point of extinction, or, if it survives, a population bottleneck

I don't see how this is relevant to our discussion. However, the relation between lack of variation and risk of extinction is, as far as I know, a longstanding, urgent, and unsolved problem in conservation biology. See, for example, Habel & Schmitt 2012 The burden of genetic diversity.

As we seem to have reached the point where we are largely in agreement, we have probably reached the end of our discussion. Thanks for talking with me!

Arno Wouters said...

Chris B. asks regarding the distinction between 'being currently under selection' and 'currently being selected for or against':

How would you explain the difference between these two phrases?

Frankly, I have no idea.

I took it for granted that 'being currently under selection' is just another word for 'currently being selected for or against' until judmarc suggested that I conflate the two.

I thought he might be right.

He indicated that the distinction is basic population genetics math. I could find it neither in the very outdated population genetics text in use when I was a student nor in more recent text books on evolutionary biology in my book case (such as the one of Futuyma), so I asked for a reference.

Can you provide one?

Arno Wouters said...
This comment has been removed by the author.
Arno Wouters said...
This comment has been removed by the author.
Arno Wouters said...

Chris, we are largely in agreement, except for one crucial point.

I wrote:

"If 'being functional' is defined in terms of selected effect 'subject to selection' is just another word for 'functional', so under that definition it can't be the case that a sequence is subject to selection because it is functional."

You answered:

Functional in this case would mean that a sequence has a biological function that affects the fitness of the organism, and is therefore subject to natural selection.

Here, I disagree: 'affecting fitness' is in an important way different from 'being selected for or against'.

A sequence, by definition, affects an organism's fitness if that organism's fitness would change if the sequence were removed (or altered). 'Affecting fitness' is, in other words, defined in terms of hypothetical fitness differences between an existing organism and a hypothetical alternative. In many cases the hypothetical alternative does not exist and in many cases it even cannot exist (if you would remove a fragment of DNA coding for an essential protein from all the cells of the organism the organism would die).

Natural selection, on the other hand, occurs only if there are real fitness differences between co-existing variants in a population.

I explain this in my Viability explanation" (1995).

Selection is also not an omnipresent entity or force that applies itself under certain circumstances.

Exactly! This is why it is so important not to confuse 'affecting fitness' with 'being selected for or against'!

Chris B said...

"Can you provide one?"

No, I can't. The phrases seem equivalent to me.

Chris B said...

Arno,

I think we largely agree here and maybe we are starting to talk past each other. I am a scientist, not a philosopher, so the utility of discussions like these, to me, is always going to be their real life biological relevance and applicability.

So I must quibble with you here:

"A sequence, by definition, affects an organism's fitness if that organism's fitness would change if the sequence were removed (or altered). 'Affecting fitness' is, in other words, defined in terms of hypothetical fitness differences between an existing organism and a hypothetical alternative."

Actually, 'affecting fitness' is defined in terms of relative fitness in the environment compared to other individuals. You cannot have differential selection unless you have variation in the population. So, for example, in real life, your example of Wollemia having a genome of entirely of junk as a consequence of the definition of junk is wrong. The only way this would be true is if Wollemia never experiences genetic mutation. Other wise, regardless of how currently little genetic variation there is in the species, there is some variation and more of it being generated all the time by mutation. In this respect Wollemia is like every other species.

Arno Wouters said...

Chris B. wrote:

Selection is also not an omnipresent entity or force that applies itself under certain circumstances.

I answered:

Exactly! This is why it is so important not to confuse 'affecting fitness' with 'being selected for or against'!

Let me amplify on this.

In the case of pigeon breeding it makes sense to say that a certain breed is under selection for traits that please the fancier because the breeder is looking for traits that please him or her even if the doves do not vary with respect to these traits. So 'being scrutinized by a person who can influence the life chances' yields a definition of 'being under selection' that doesn't require actual existing variation.

However, this definition doesn't work for natural selection because, as you say, there is no scrutinizing force or agent. All we have are differences in fitness between co-existing variants. So how could we talk about a DNA-fragment (or any other trait) being under natural selection if there is no variation with respect to that trait between co-existing variants and no force or agent causing the selection?

I am not saying that natural selection is a force or agent causing selection. On the contrary, I have written a short series of blogposts in my native language (Dutch) to explain that natural selection shouldn't be seen in this way!

What I am saying is that for the moment I don't understand how to make sense of the notion of 'begin currently under selection' in absence of both variation and a scrutinizing force or agent.

I am open to the possibility that there is a way to do so and for that reason I asked for a reference to a basic population genetics math text.

past selection leaves its mark on sequences that have historically been under selection

Sure, I am aware that past purifying selection can be detected from sequence conservation and as I said, I think that sequence conservation is good evidence for function. What I am saying is that if you think that having a function explains selection it makes no sense to define function in terms of past selection.

I am repeating myself, which, I think, is a sign that it might be better to bring this discussion to an end by agreeing that we disagree.

Arno Wouters said...

Chris, my previous post was written and posted before your comments of Saturday, April 09, 2016 10:13:00 AM and Saturday, April 09, 2016 10:21:00 AM appeared in my reader.

Arno Wouters said...

the utility of discussions like these, to me, is always going to be their real life biological relevance and applicability

Fair enough!

Actually, 'affecting fitness' is defined in terms of relative fitness in the environment compared to other individuals. You cannot have differential selection unless you have variation in the population.

Oops, you are right that population geneticists usually define 'affecting fitness' in terms of relative fitness. I should have said that many functional biologists talk about fitness contribution in terms of hypothetical fitness differences between existing organisms and hypothetical alternatives.

I belief that one of the ways in which my Four notions might be relevant to biologists is that it draws attention to the different ways in which the term 'function' is used in different parts of biology and explores some of the ways in which these different notions are related.

For that reason I feel embarrassed that I forget to add that qualification that what I was talking about the way many functional biologists talk about fitness contribution.

However, this omission doesn't affect my point (with which you seem to agree) that something can be functional in the sense that it contributes to absolute reproductive succes without currently being selected for or against.

So, for example, in real life, your example of Wollemia having a genome of entirely junk as a consequence of the definition of junk is wrong. The only way this would be true is if Wollemia never experiences genetic mutation. Other wise, regardless of how currently little genetic variation there is in the species, there is some variation and more of it being generated all the time by mutation. In this respect Wollemia is like every other species.

Hmmm. I thought it clear that I was deliberately exaggerating in order to get my point through, when I said that under a definition of junk that implies that a DNA segment is functional only if it is currently under selection (in the sense of being selected for) the entire genome of Wollemia would be junk ....

Given your explanation I obviously wasn't.

Perhaps, I should be aware that this way of bringing an argument is not understood outside of philosophy.

I'll try to rephrase my argument without such exaggeration:

My point is that the term 'function' as it is ordinarily used in the context of the junk DNA debate is not function as defined by Doolittle and Georgi ("segments of DNA are functional if they were historically selected for in the process of evolution and are maintained by purifying selection today") but function as biological role (e.g. as contributing to an organism's ability to survive and reproduce).

In order to do so I try to envisage situations in which the proportion of DNA-segments that result in fitness differences between co-existing organisms in the population is smaller than the part of the genome that you and I (and other participants in this discussion) would call 'functional'.

I believe this is the case with Wollemia. Do you agree?

More generally, would it matter to your verdict about function how much variation is currently going on in the population?

If the answer to my first question (do you agree) is 'yes' and/or the answer to the second (does the amount of current variation matter) 'no', wouldn't this indicate that your implicit notion of function cannot be the Doolittle and Georgi's notion of selected function as that notion implies that a functional trait is currently being selected for?

Arno Wouters said...

Another way in which I hope my function papers (especially my Biology's Functional Perspective) are relevant outside philosophy lies in my explanation of how it is possible to talk about function in absence of teleology and sentience. This is probably not relevant to biological research, but I do think it might help science teachers and other people who are dealing with intuitive teleologists.

Unknown said...

My earlier link is to NSF- but that site is down.

Here is another link where the findings are covered-
http://www.livescience.com/45205-data-dont-back-up-darwin-in-algae-study-nsf-bts.html

“One of Charles Darwin's lesser-known hypotheses posits that closely related species will compete for food and other resources more strongly with one another than with distant relatives, because they occupy similar ecological niches. Most biologists long have accepted this to be true.
Thus, three researchers were more than a little shaken to find that their experiments on fresh water green algae failed to support Darwin’s theory— at least in one case.”

If Eric is correct about the state of the scientific society, then I would think these researchers would have been ‘punished’ or put in place by now. Have they been? I don’t know… I’m hoping Eric will give us a report.

Chris B said...

Arno,
Thanks for your clarifications. I think I understand your reasoning better here.

"In order to do so I try to envisage situations in which the proportion of DNA-segments that result in fitness differences between co-existing organisms in the population is smaller than the part of the genome that you and I (and other participants in this discussion) would call 'functional'.

I believe this is the case with Wollemia. Do you agree?"

Yes, I agree, in fact I think this is true for every species on Earth. Every functional part of the genome is not under observable selection at all times, yet is still functional.

"More generally, would it matter to your verdict about function how much variation is currently going on in the population?"

No, functional parts of the genome will be functional regardless of the level of variation at any given moment in time.

"If the answer to my first question (do you agree) is 'yes' and/or the answer to the second (does the amount of current variation matter) 'no', wouldn't this indicate that your implicit notion of function cannot be the Doolittle and Georgi's notion of selected function as that notion implies that a functional trait is currently being selected for?"

Yes, I would say that is a fair and logical conclusion. I would include not only traits being selected for, but parts of the genome that show historical signs of selection. Those are indicators of function as well.

I like your definition of function as biological role, as long as you define it as you did above, "as contributing to an organism's ability to survive and reproduce)."
The danger without a rigorous definition of biological role is that one may define it as, say, transcriptional activity, which brought us the ENCODE fiasco in the first place. As critics of ENCODE have pointed out, one could have defined biological role as replication, rendering the genome 100% functional.

Jmac said...

Thank you Chris B. It is really "refreshing" to read your shit. It really makes me feel good when a bunch of retards try to change the way the things are... If I were the ID, I would make you so humiliated you would not be able to cry.

One thing I know about the ID, the ID has patience beyond his design, otherwise ID would have killed all his enemies and possibly me...

SRM said...

If I were the ID, I would make you so humiliated you would not be able to cry.

and that would be a rather rare show of restraint for this loving christian god

One thing I know about the ID, the ID has patience beyond his design....

Could you tell us more about the Intelligent Designer's design? I'm assuming it must be by some other intelligent designer. This should be interesting. But it won't be, because that line of inquiry does not compute with Eric et al.

Jmac said...

The final summary I missed yesterday:

1.The ENCODE people don't seem to care if there is a consensus as to WHICH EVOLUTIONARY THEORY SHOULD BE TAUGHT in schools next semester. All the ENCODE seem to care about is what kind of deseases, like cancer, can possibly be prevented or cured by their research results.

Am I wrong about this whole thing? Am I delusional? It's actually my best joke!

Well, let's analyze the evolutionary side:

For over 50 years, since the unknown function of the majority of human genome was called and accepted by evolutionists as "junk DNA", many loved the notion and accepted it with pride due to their own or their 'brothers' predictions.

This theory seem to have changed now. It seems opposite to the evolutionary predictions, although people like Larry and his comrades protest; Junk DNA is still junk to them.

Here is the main point: If scientists following up on the ENCODE findings, find new, specific functions to the previously assumed part of genome as "junk", how should the opposers of this notion react? Should they be happy or sad? You'd be the judge...

The truth is that evolutionist have insisted and some of them still do, not to investigate the so-called junk DNA, and put pressure on those what objected to it and continued their research.

Today, there is no doubt about who was right and who was wrong. Unfortunately, the defeated side-the evolutionists continues to claim that they are right and the the ENCODE and their supporters are wrong.

How So?
What is their scientific evidence? Is their belief as good as scientific evidence? If yes, why?

Jmac said...

So... the only conclusion can be drawn from this discussion is that the evolutionists, like the ones mentioned by me specifically above, must have hindered the scientific progress into the research of their "imaginary junk DNA" that now turns out to hold the insights into not only many human deseases; that can possibly be prevented or even cured, if the ENCODE findings are taken seriously unlike the evolutionist have endorsed it be...

Chris B said...

So no arguments, just insults. Thanks, Eric. How not refreshing. I really didn't think you would come through with any facts or good arguments, but hope springs eternal.

"If I were the ID, I would make you so humiliated you would not be able to cry. "

I suppose incoherent babble could make me cry if I was subjected to it long enough. Your point?

Chris B said...

After all the attempts to explain reality to you, you still manage to get every single point wrong again, Eric.

I think you are right, Eric, you are delusional.

Chris B said...

Eric,

"It really makes me feel good when a bunch of retards try to change the way the things are.."

Why do ID/creationists project their inadequacies onto others so much? Are you even aware of it?

judmarc said...

a bunch of retards

I want to hear again about how your religion does such a great job of showing you the correct moral path.

Jmac said...

Is it a delusion, when the pillars of science choose to vindicate their own egos by ignoring the scientific evidence?
I'm not talking about ENCODE anymore that evolutionists hate. I'm talking about the follow up papers that assign specific functions to the regions of DNA that evolutionist doomed as junk and discouraged other scientists to investigate it. Today... it is a day of vindication.... unfortunately, it is not for evolutionists...

Chris B said...

Yes, Eric, you are delusional, because what you just said there is not part of reality. Or maybe you're just lying.

Corneel said...

@Arno Wouters

My main objection is the presence of multi-level selection. I tried to illustrate this with the example of selfish elements, which are being subject to selection while being deleterious for the host. Another example would be altruism genes (e.g. green-beard genes). When you take functional to mean important in survival in reproduction, You automatically default to the level of the organism. Of course, the organism is often the natural unit (especially for biomedical research), but genes do not always evolve in the service of the organism, nor is the organism always easy to distinguish in biology.
Genes evolve by virtue of their transmission efficiency. This may be due to a beneficial contribution to the organism, but this is not necessary. Hence the definition of function as selected effect is not meant to facilitate a "lazy form of biology" but to alleviate problems with other definitions that always seem privilege the organismal perspective.

Arno Wouters said...

Thanks for your clarification Corneel!

Although multi-level selection might be part of the explanation of how junk DNA (defined as DNA that doesn't have a function as biological role) came into being and why it is not weeded out by selection, I don't see how multi-level selection is relevant to my point that the relevant notion of function in the junk DNA debate is function as biological role (in the sense of contributing to an organism's ability to survive and reproduce).

Let's have a look at Doolittle's Is junk DNA bunk?

The paper starts with the main argument for the view that most of our genome is junk, namely "the perennial problem of C-value". The gist of this argument is that given the lack of correlation between amount of DNA and organismal complexity a large part of the DNA of organisms that have a relatively large amount of DNA does not contribute to their survival and reproduction. I quote (p. 5295):

If we do not think of this additional or “excess” DNA, so manifest through comparisons between and within biological groups, as junk (irrelevant if not frankly detrimental to the survival and reproduction of the organism bearing it), how then are we to think of it?.

So the conclusion of the C-value argument as presented by Doolittle, clearly is excess DNA doesn't have a function as biological role.

Next, Doolittle discusses how the existence of junk DNA fits into evolutionary theory. He seems to favor Gregory's view, which Doolittle summarizes as follows

A balance between organism-level selection on nuclear structure and cell size, cell division times and developmental rate, selfish genome-level selection favoring replicative expansion, and (as discussed below) supraorganismal (clade-level) selective processes—as well as drift— must all be taken into account.

In other words: the existence of DNA that doesn't have a function as biological role can be explained a combination of multilevel selection and drift.

(to be continued)

Arno Wouters said...

(continuation)

After have explained clearly that the C-value argument shows lack of function as biological role, Doolittle suddenly and very confusingly introduces another definition of function:

Most philosophers of biology, and likely, most practicing biologists when pressed, would endorse some form of the selected effect (SE) definition of function (28–30)

Unfortunately, he doesn't provide evidence for this empirical claim about how two groups of academics verbally define 'function'.

More importantly, Doolittle doesn't explain how the statement that "Most philosophers of biology, and likely, most practicing biologists when pressed, would endorse some form of the selected effect (SE) definition of function" is supposed to support his view that (a) the term function traditionally refers to function as selected effect, and (b) the notion of function as selected effect should be preferred over function as biological role in the context of the junk-DNA debate.

Doolittle's redefinition of function in terms of selected effect is, furthermore, very confusing.

One reason is that this definition is at odds with the way in which Doolittle presents the C-value argument. As I have argued above, the conclusion of this argument (at least as Doolittle presents it) is: excess DNA doesn't have a function as biological role. Because this phenomenon can be partly explained as the result of multilevel selection it, follows that the argument doesn't show that excess DNA doesn't have a function as selected effect.

So, if junk is defined in terms of selected effect, the C-value argument doesn't show that excess DNA is junk.

Another reason is the teleological character of the notion of function as selected effect. This notion was developed by Karin Neander and by Ruth Millikan in order to accommodate teleological notions such as 'the function of', 'what it is for', 'in order to' and 'for the purpose of' into a Darwinian framework. To do so they redefine these notions in terms of 'what it was selected for in the past and explains its current existence'. A clever trick, ably used to solve a number of longstanding problems in the philosophy of mind and language. Great!

However, the teleological notion of function as past selected effect deprives function claims of their explanatory force, as I explained in a previous comment. Hence, function as selected effect cannot be used to explain selection.

So although, Neander and Millikan offer an account of how it is possible to talk about purposes etc. within a Darwinian framework, their account doesn't legitimize appeal to such notions in explanations of how things came to be the way they are.

This lack of explanatory force of functions as selected effect is not relevant to the philosophical problems addressed by Neander and Millikan, but it is a very serious problem for the application of the notion of function as selected effect in biology.

Doolittle continues:

Selected effect is the form of teleological explanation allowed, indeed required, by Darwinian theory

Yes, you read it well, that's what Doolittle says: "teleological explanation is allowed, indeed required, by Darwinian theory"!!!

A clear example of the kind of confusion instigated by defining function in terms of selected effect!

I didn't say and I don't think that the definition of function as selected effect is 'meant to facilitate lazy biology'. Rather, I think that this definition de facto and all too easily facilitates lazy biology and teleological ways of thinking.

I refer, once more, to my Biology's Functional Perspective for further explanation._

(The End)

Rolf Aalberg said...

Who's right and who's not? All I can say it looks like Eric needs a place to hang his hat.

Is ENCODE his version of Noah's Ark?

Jmac said...

You see Chris, most people in the evolution business-(because that is exactly what it is)-who really care about their reputation have adjusted their views, at least a bit, because of ENCODE. But some did not ( I can guess why, I think) and that includes Larry Moran, Dan Graur, PZ. Myers and the upcoming star of the evolution business Georogi Marinov.

Except for the later, they couldn't care less if they are ever proved wrong in few months or years.

I'm no gambler, but in 5 years all of them will have to swallow the truth. Georgi will have to swallow razors but he will do it and he will move on. That's the nature of this business he is in. It has nothing to do with science. It never did and it never will.

«Oldest ‹Older   1 – 200 of 206   Newer› Newest»