Luskin: Dual-Coding Genes “Nearly Impossible by Chance” — How Would Francisco Ayala Respond?

| 47 Comments

On Uncommon Descent Luskin asks Ayala the following question:

How would dual coding genes, which are nearly impossible to arise by chance, evolve via Darwinian processes?

Luskin “argues” that

Leading evolutionary biologist Francisco Ayala recently wrote in Proceedings for the National Academy of Sciences (PNAS) that “Chance is an integral part of the evolutionary process.” Ayala then explained why he thinks Darwinian evolution is right and ID is wrong: “Biological evolution differs from a painting or an artifact in that it is not the outcome of preconceived design. The design of organisms is not intelligent but imperfect and, at times, outright dysfunctional.” (“Darwin’s greatest discovery: Design without designer,” PNAS, 104:8567–8573 (May 15, 2007), emphasis added.) This questionable standard and conclusion is Ayala’s punchline against ID.

Ignoring for a moment the empty rhetoric of Luskin, let’s explore how Ayala may answer the question. Oh wait…

Ayala already answered the question

Chance is an integral part of the evolutionary process. The mutations that yield the hereditary variations available to natural selection arise at random, independently of whether they are beneficial or harmful to their carriers. But this random process (as well as others that come to play in the great theatre of life) is counteracted by natural selection, which preserves what is useful and eliminates the harmful. Without mutation, evolution could not happen because there would be no variations that could be differentially conveyed from one to another generation. But without natural selection, the mutation process would yield disorganization and extinction because most mutations are disadvantageous.

Of course, the answer is also found in the paper which outlines the dual coding genes.

Yet in cases of tightly coexpressed interacting proteins, dual coding may be advantageous. Here we show that although dual coding is nearly impossible by chance, a number of human transcripts contain overlapping coding regions.

Seems that there may indeed be advantages to dual coding, in other words, random variation AND selection, which are exactly the mechanisms of Darwinian evolution.

Now that Ayala has answered Luskin’s question, perhaps Luskin can enlighten us how ID explains dual coding? Oops, I forgot, ID does not deal with such ‘pathetic’ questions, to paraphrase ID defender Dembski.

And people wonder why ID remains scientifically irrelevant? It is doomed to remain so by its own foundation in ignorance.

Now I understand why Luskin failed to link to the article in PNAS, Ayala explains it all.

47 Comments

Seems that Luskin’s portrayal of the paper which argues that it is nearly impossible to arise by chance is also flawed, the paper in question states

. Only 0.3% of simulated alignments preserved ARFs with 500 or more nucleotides (Figure S2). Thus, both simulations suggest that only a negligible amount of random dual-coding regions will reach 500 bp, and we set this length as the threshold for defining ARFs in orthologous coding regions.

While surely a small number, this hardly meets the requirements set by ID of 1 in 10^150…

Wait, what? Overlapping genes are ‘magic’ to IDers? What? Am I reading this right? Topic covered on page… 40-ish of the Intro Genetics book Im looking at right now? What?

I read that PLoS Comput Biol article when it came out, but I am not sure I understand their methods (too mathematical for me). If I interpret their conclusions correctly, they say that by chance one would expect an ARF of >500 bp to overlap an existing ORF in 0.1-0.3% of cases (depending on assumptions). Based on a 21,000-gene genome, that would mean 20-60 alternative reading frames by chance alone. Or am I missing something?

If that’s the case, the occurrence of each ARF is indeed rare, but several should pop up regularly as genomes evolve, and those with selectable functions of course will be maintained.

By the way, here is Chung’s paper, for anyone interested (free access).

*blink*

Sooo… HIV is evidence for ID Creationists god once again.

Id like to thank their god personally for my career. Without Him, Id be forced to spend the day playing with my dog and eating ice cream. Does anyone know His phone number?

Or is it just dual coding in humans thats magic? Like how drug resistance in HIV is normal to Behe, but drug resistant malaria is magic?

I think the authors make a mistake with the ‘virtually impossible’ comment. 0.3% is not virtually impossible. It only gives fuel for the IDers. Also the ‘virtually impossible’ only refers to generating the sequence by randomization. It has nothing to do with evolutionary processes (ie an ARF can start out smaller, than grow larger in size if it offers evolutionary advantage)

Sorry to go off-topic, but something very strange seems to have happened to the “Biblical Inerrancy vs. Physical Evidence” post: a copy of the entire thread seems to have been embedded within one of the comments, and the screen-refresh rate seems to have slowed drastically. At least that’s how it looks from my PC. Y’all may want to look into that.

Or it could just be me. I’ll check back later…

Andrea Bottaro Wrote:

Based on a 21,000-gene genome, that would mean 20-60 alternative reading frames by chance alone.

And amazingly enough, they found 40 alternative reading frames - smack dab in the middle of what is predicted by the chance hypothesis!

Of course, if you read the paper - thanks for the link, btw - the authors are actually arguing that if you find a potential ARF of over 500 base pairs, it is likely that it has been selected for (functional), rather than being a recent or neutral mutation. At least, that is my amateur take on the paper, especially in light of the first two sentences of their conclusion:

Maintenance of dual-coding regions is evolutionarily costly and their occurrence by chance is statistically improbable. Therefore, an ARF that is conserved in multiple species is highly likely to be functional.

In other words, it’s not that it’s unlikely to arise at some point, it’s that it’s unlikely to both arise and stick around unless it is beneficial. If you find a potential ARF in multiple species, it’s a damn good bet it’s actually doing something independent of the ORF.

Luskin wrote:

The design of organisms is not intelligent but imperfect and, at times, outright dysfunctional.” … This questionable standard and conclusion is Ayala’s punchline against ID.

The reason he writes this is because we should not expect the designer to just design perfect stuff. This, of course, follows from ID inability to say anything about the designer. And yet, as most people already know, Luskin is VERY fond of claiming that ID predicts that there should be no junk DNA.[Dr Evil]Rrrriiiiight![/Dr Evil]

Hypocrite

These IDiots just can’t get it through their thick skulls that “nearly impossible” = “possible” = “likely for sufficiently large numbers of events”. And when determining how many events are sufficient, you have to get the math right, as they never do, because they treat all the events as independent, when the key insight of Darwin is that they are dependent, due to selection. When, over and over again, the question is about the climb up Mt. Improbable, you would hope that even someone as dense as Luskin could grasp the concept of “asked and answered”.

I think the authors make a mistake with the ‘virtually impossible’ comment. 0.3% is not virtually impossible.

I think you make a mistake by posting here without engaging your brain or eyeballs – no one but you used the phrase “virtually impossible”.

Also the ‘virtually impossible’ only refers to generating the sequence by randomization

Gee, I wonder if that has anything to do with why Ayala qualified it as “nearly impossible by chance”, and noted that “this random process … is counteracted by natural selection”.

So, sure, if you ignore what Ayala wrote and twist his words around, you can pretend that the “fuel” for IDers doesn’t come out of their own puny and dishonest brains and blame the victim of quote mining for having his quotes mined. It’s an all too common but rather sickening approach.

The paper did use the terms nearly and virtually

Here we show that although dual coding is nearly impossible by chance, a number of human transcripts contain overlapping coding regions.

Dual Coding Is Virtually Impossible by Chance

In this light Popper’s G’s statement seems particularly ironic

I think you make a mistake by posting here without engaging your brain or eyeballs — no one but you used the phrase “virtually impossible”.

:wink:

Dual Coding Is Virtually Impossible by Chance

Is this a direct quote? In any case, as I said, the “by chance” part is critical, and LPalmer’s omission of it is quote mining – the results of evolution are not achieved “by chance”, they are achieved by selection. It’s amusing that, on the TV poker shows, they point out repeatedly that, despite the luck of the draw, the same faces show up over and over in the final round – they grasp the difference between random processes and selective but noisy processes.

Hey, when is Flint going to come along and dismissively pontificate that

Luskin is only doing what lawyers do - building a case in support of the desired outcome, without any compelling need to consult reality if it doesn’t contribute what he needs. Let me pick the jury, and Luskin will win every time.

Of course, aside from that fact such behavior is not universal among lawyers in all instances – his claim was particularly slanderous when posted in lawyer Timothy Sandefur’s thread – Luskin isn’t only doing that; he is also making erroneous claims that warrant rebuttal. And, for what it’s worth, Flint’s later claim that Luskin is neither stupid nor ignorant is not supported by the facts. He, stupid, he’s ignorant, and he’s intellectually dishonest.

Dual Coding Is Virtually Impossible by Chance

I finally found this in Chung’s paper. I don’t think there’s anything “ironic” about my comment when you turkeys can’t be bothered to say what you’re talking about. Notably, according to your piece, Luskin addressed Ayala, not Chung. Perhaps he addressed Chung too, but you didn’t bother to provide a link to Luskin’s comments.

In any case, Chung’s claim is correct; it is virtually impossible that the 40, very conservatively, conserved ARF’s are the result of chance.

Luskin, citing Chung et al, Wrote:

Here we show that although dual coding is nearly impossible by chance, a number of human transcripts contain overlapping coding regions. [Luskin’s emphasis.]

Luskin digs his own grave, as the paper shows evolution at work.

In other news, priest kicks dog.

LPalmer Wrote:

I think the authors make a mistake with the ‘virtually impossible’ comment.

Well, that is a header for a chapter where they are motivated to establish that their method picks out ARF’s reliably. “To reliably find new dual-coding genes, we must determine how likely they are to occur by chance.”

PvM Wrote:

On Uncommon Descent Luskin asks

Not at UD, but at Luskin’s usual haunt Evolution News.

Ok, you did give the link to Luskin’s ‘arguments’ in the comments, and Luskin was addressing the Chung paper. So LPalmer is right in using the word “virtually” (but not right in his quote mining that omitted the crucial “by chance”), I’m wrong that no one said it, and “ironic” is a rather polite term for my major screwup. My sincere apologies for being such a dufus.

But on the substance – “virtually impossible” is false, but “virtually impossible by chance” is true. .3% is not the chance in question – the chance in question is that of there being 40 – very conservatively – conserved ARF’s. Rather than providing fuel to IDists, the paper is presenting a statistical argument that they are a result of selection. When they said “virtually impossible by chance”, that’s really what they meant – and whether they are right or wrong, it doesn’t give a whit of support to ID arguments.

Torbjörn’s and my postings crossed.

Well, that is a header for a chapter where they are motivated to establish that their method picks out ARF’s reliably.

Exactly. What is a mistake is to ask scientists to avoid certain claims or hypotheses because some cdesign proponentist might take it out of context and misconstrue it, and build a fallacious argument around it. It’s doubly mistaken when “we” help them out by doing so ourselves.

Upon quickly browsing it I find Ayala’s paper quite good, though it covers well trodden ground and is framed as explaining apparent design.

Luskin however has no real answer. Which is why he must take an example of evolution and try to make it out that it goes against Ayala’s description of “dysfunctional” design.

Which traces of course here would be the unfixated ARF’s that the paper didn’t really look for. However, they rejected 26 ARF’s that were only conserved over two of three species out of 66. So there is Luskin’s answer, ARF’s appears as dysfunctional as any other evolutionary result.

PG Wrote:

you didn’t bother to provide a link to Luskin’s comments.

Comment #183776.

PG Wrote:

Torbjörn’s and my postings crossed.

Seems this time my and Popper’s Ghost’s postings crossed. Sorry, you already noted the link.

Luskin Wrote:

We mortals are easily impressed by palindromes — words or phrases that have the same spelling forwards and backwards. One meaning is gained when you start with one letter of the first word, and then an entirely different meaning is understood when you start reading with the second letter of the first word. Such a sentence would be most impressive, but what if such “sentences” existed in our DNA?

Well, it shouldn’t be impressive at all, because DNA was created through a mechanism, and mechanismss are quite good at producing that sort of thing. If there is such a sentence, it wouldn’t be much work to produce a program to generate it from a list of words. Going back to palindromes, it’s hard for people to come up with really long ones, but here’s one with 17259 words that was mechanically produced:

http://norvig.com/pal2txt.html

Luskin would have an even harder time with the article by Michael Lynch in the same PNAS supplement.

PvM Wrote:

Now that Ayala has answered Luskin’s question, perhaps Luskin can enlighten us how ID explains dual coding? Oops, I forgot, ID does not deal with such ‘pathetic’ questions, to paraphrase ID defender Dembski.

Actually Ayala did not answer Luskin’s question, because nothing short of a nanosecond-by-nanosecond account of the whereabouts of every atom would stop him from saying “that pathetic level of detail doesn’t answer my question.”

While evolution answers that question with ever-increasing detail, and provides testable predictions along the way, and ID doesn’t even try (and even evades the most “basic what happened when” questions), evolution is still at a disadvantage as long as most of the public is sold on ID’s double standard of evidence.

Maybe if we take an occasional break from “ID ‘is’ creationism” and “ID sneaks in God,” and show people what a slick game ID is, will they begin to see how they are being scammed.

If you click the link in Comment 183867, make sure to scroll to the first of 3 posts. The part about how IDers know (privately at least) that it must be evolution (if not their “Darwinism” caricature), and how they must agree (privately at least) that any “discontinuities” do not disrupt the “biological continuity” of common descent is more compelling than ever in light of their approval of Michael Behe’s new book.

OK, I guess in all the bickering my point on the paper got lost. Does anyone understand from their methods section whether the algorithm does/should correct for the number of genes?

From my understanding, they take an ORF, check if it has any ARF longer than some arbitrary value, than change codons around on the ORF in simulated alignments and see whether the ARF is still there. If it’s there in only a rare combination of codons on the ORF that allows for the ARF, they consider it a statistically significant (“real”) ARF.

What I am not sure of, does this process need to correct for the fact that there are tens of thousands of ORFs in the genome, and the frequency by which these can potentially encode ARFs? Or should it only consider ORFs on an individual basis? I think this is the key to understand the “virtually impossible” statement.

Syntax Error: not well-formed (invalid token) at line 10, column 153, byte 987 at /usr/local/lib/perl5/site_perl/5.12.3/mach/XML/Parser.pm line 187

My use of the 14159 number of genes is not going to be correct. Really one should determine the number of ORFs > 500.

Still it would be nice to know the statistical probability of observing 149 genes, or the statistical probablity of observing none.

And, for what it’s worth, Flint’s later claim that Luskin is neither stupid nor ignorant is not supported by the facts. He, stupid, he’s ignorant, and he’s intellectually dishonest.

I think we can safely conclude, on the facts, that Luskin is using rhetorical techniques PG finds offensive, in an effort to support conclusions PG knows are incorrect, out of motivations PG does not share. Intellectually dishonest is problematic. But “stupid and ignorant” here, on the facts (I love that phrase for personal opinions!) seems to mean “disagrees with PG”.

A little perspective on the DI may be called for. Imagine that you are being paid to distort known facts to support conclusions they do not support (and indeed, refute). Imagine that your religious faith tells you that your conclusions MUST be correct, and if they’re apparently refuted by all known observation, then something is wrong with the observations and/or the interpretation of those observations. It MUST be, since your god told you you’re right beyond question.

Now, whether or not the sincere effort to bend reality to ratify even the wackiest religious belief is “intellectually dishonest”, does the absurdity of your religious doctrines ipso facto render you stupid or ignorant of what you’re trying so hard to fit to your requirements? In my opinion, the DI folks are generally capable, intelligent, diligent people harnessed to incorrigible delusions. Stupid, ignorant liars would present little need to combat. But these people are smart, educated, and insane. A much tougher nut, which explains all these threads. Just because a soldier is fighting for the enemy doesn’t mean he’s unarmed.

Syntax Error: not well-formed (invalid token) at line 2, column 393, byte 526 at /usr/local/lib/perl5/site_perl/5.12.3/mach/XML/Parser.pm line 187

Andrea Bottaro Wrote:

OK, I guess in all the bickering my point on the paper got lost. Does anyone understand from their methods section whether the algorithm does/should correct for the number of genes?

From my understanding, they take an ORF, check if it has any ARF longer than some arbitrary value, than change codons around on the ORF in simulated alignments and see whether the ARF is still there. If it’s there in only a rare combination of codons on the ORF that allows for the ARF, they consider it a statistically significant (“real”) ARF.

What I am not sure of, does this process need to correct for the fact that there are tens of thousands of ORFs in the genome, and the frequency by which these can potentially encode ARFs? Or should it only consider ORFs on an individual basis? I think this is the key to understand the “virtually impossible” statement.

Some good points Andrea. I think we have made one mistake in our interpretation of the data. The .3% number seems to be the percentage of alignments of 5000 columns of codons that have Arfs. I don’t know how this relates to number of genes (which would mean my numbers in my previous post are going to be wrong)

I don’t think they do any correction for number of genes in their test of ARFs. So I assume there are going to be some false positives in their 40 genes, but I don’t think that matters too much.

one would think that at least a couple of IDers would have seen Princess Bride (given their propensity for Vinzini like logic). Just as Miracle Max knows the difference between mostly dead and dead, just as the there is difference between nearly impossible and impossible.

Miracle Max:Hoo hoo hoo! Look who knows so much, heh? Well, it just so happens that your friend here is only mostly dead. There’s a big difference between mostly dead and all dead. Please, open his mouth. Now, mostly dead is slightly alive. Now, all dead…well, with all dead, there’s usually only one thing you can do.

If I’m understanding correctly what “dual coding” means, it sounds like something that a competent bioengineer would avoid, unless it there was some function that couldn’t be produced without it.

Henry

If I’m understanding correctly what “dual coding” means, it sounds like something that a competent bioengineer would avoid, unless it there was some function that couldn’t be produced without it.

ID makes no statement about who the Designer might be, nor what level of skillz the Designer might possess.

We can only infer that:

1. The designer is so brilliant that we can’t determine how s/h/it designed things.

2. But, despite this, the designer’s *bleeping* stupid, making all sorts of beginner-designer mistakes.

3. This is why the designer must be worshipped.

Sentence 1, “Neither of Luskin’s claims are true.”

Sentence 2, “Either of Luskin’s claims are true.”

Wow, is Luskin impressed yet?

“aseptic tanks should be used for pharmaceutical solvents”. “septic tanks should be used for pharmaceutical solvents”.

“skipper was thrown overboard” “kipper was thrown overboard”

It is obvious that Francisco Ayala answered what ever Luskin was questioning/implying anyway. That the answer was already in the essay and Luskin was not able to understand it is not a surprise either. Luskin cannot write coherently how he pretend to argue logically with the likes of Ayala who is light years ahead of him? Oh, I forgot Luskin must have received his knowledge through divine revelation and thus he knows everything.… NOT.

“This is an interesting game.” “His is an interesting game.”

“Not going to use acronyms?” “OT going to use acronyms?”

Cool! So the question was “How would dual coded sections of DNA evolve?” boils down to “By random mutation and natural selection.”

Fantastic. Ten out of ten for accuracy, and minus several million for usefulness. Do you work for Microsoft?

But seriously, folks .… the point Luskin was making was that the chances of random mutations generating something that dual codes is very small. And the reason he made this point was because each random evolutionary step along the way has to convey an advantage for natural selection to work with. It’s easy to see that a dual-coding section of DNA conveys a significant advantage. What isn’t easy to see is how you get there from a non-dual-coding piece of DNA.

You have a choice. You can either continue to pat yourselves on the back over another mythical (not even pyrrhic) victory over the forces of darkness, or you can say, “Actually, to be honest, we haven’t a clue, and we don’t have a handle on the numbers involved.” Or you could suggest something a bit more scientific in terms of process and calculation.

Incidentally, PvM, a fully specified sequence of 500 BP has an improbability of about 1 in 10^300. It may not be the same as 1 in 10^150 - but it does acknowledge a limit.

… and another thing.

“I saw a peacock with a fiery tail. I saw a blazing comet drop down hail. I saw a cloud with ivy circles round. I saw a sturdy oak creep on the ground. I saw a pismire swallow up a whale. I saw a raging sea brim full of ale. I saw a Venice glass sixteen foot deep. I saw a well full of men’s tears that weep. I saw their eyes all in a flame of fire. I saw a house as big as the moon and higher. I saw the sun even in the midst of night. I saw the man that saw this wondrous sight. - Anon.”

There’s an example of of a dual-coding text - and the two codes are properly displaced, not overlapping. That’s what’s clever about it. Is that the product of an unintelligent process, I wonder?

But that’s not a dual-coding stretch of text, yet. Can anyone think of a string of text - over, let’s say, 20 characters, that conveys two different, unrelated meanings depending upon where you start? I daresay somebody can. But it would take careful thought and planning.

It’s easy to see that a dual-coding section of DNA conveys a significant advantage.

Are you kidding? It’s a distinct disadvantage, since it would hinder adapting one of the codes to a change in environment without messing up the other one.

Exiled from GROGGS wrote:

“And the reason he made this point was because each random evolutionary step along the way has to convey an advantage for natural selection to work with. It’s easy to see that a dual-coding section of DNA conveys a significant advantage. What isn’t easy to see is how you get there from a non-dual-coding piece of DNA.”

As Henry J pointed out, dual coding is probably not an advantage, unless extreme economy is selected for, which it definately is not in eukaryotes.

Also, it does not have to convey an advantage at every step along the way, it just has to be neutral or only slightly deleterious at every step. The function of the first reading frame would maintain the sequence, while the second reading frame would be constrained by the requirements of the first but could still change considerably over time. If this type of thing were truly adaptive then the genetic code would probably not be read in a non-overlapping fashion.

What I mean is that it’s easy to see that a dual-coded piece of DNA conveys an advantage compared to a single-coded piece if the second function that it encodes would otherwise have been absent. From a redundancy POV, you are right - it would make more sense for the two functions to be separate.

Exiled from GROGGS,

Agreed. So once again, there is not theoretical reason why such a system could not arise through random mutation and natural selection and no real reason for an intelligent agent to design the system this way from scratch. Makes perfect sense from an evolutionary perspective but not from an ID perspective.

But “stupid and ignorant” here, on the facts (I love that phrase for personal opinions!) seems to mean “disagrees with PG”.

No, by “the facts” I mean the actual facts concerning Luskin’s pursuits and the arguments he makes, as opposed to the factless gas emanating from your nether regions on which you base your claim that Luskin isn’t stupid and ignorant. Being an educated lawyer, and even clever and devious, does not guarantee that one is neither stupid nor ignorant, a fact that you seem unable to reach through your elitist fog.

But seriously, folks …

Seriously, did you read any of the previous posts, here, moron?

Sentence 1, “Neither of Luskin’s claims are true.”

Sentence 2, “Either of Luskin’s claims are true.”

Hey, this is fun:

“Atheism is nonsense.” “Theism is nonsense.”

“Abuses are associated with racism.” “Buses are associated with racism.”

“Covert violations of the law are common in the Bush/Cheney administration.” “Overt violations of the law are common in the Bush/Cheney administration.”

“Gastronomical discoveries awaited him.” “Astronomical discoveries awaited him.”

“Grapes were on her mind.” “Rapes were on her mind.”

“Masses rallied.” “Asses rallied.”

“Pouting gays should be avoided.” “Outing gays should be avoided.”

“Spam had taken over his mailbox.” “Pam had taken over his mailbox.”

“Straps held the bear.” “Traps held the bear.”

“Terror is a common element of Bush’s speeches.” “Error is a common element of Bush’s speeches.”

“Weighty thoughts crossed my mind.” “Eighty thoughts crossed my mind.”

“Wellington was a famous duke.” “Ellington was a famous duke.”

“Wheels turned.” “Heels turned.”

“Wrappers were everywhere.” “Rappers were everywhere.”

“Yearly warnings were ignored.” “Early warnings were ignored.”

“Your side will win.” “Our side will win.”

“Zany ideas should be rejected.” “Any ideas should be rejected.”

etc.

About this Entry

This page contains a single entry by PvM published on June 19, 2007 11:16 AM.

Let’s Talk Junk was the previous entry in this blog.

Tangled Bank #82 is the next entry in this blog.

Find recent content on the main index or look in the archives to find all content.

Categories

Archives

Author Archives

Powered by Movable Type 4.361

Site Meter