I expressed some skepticism about that book before I read it: Philip Ball's new book: "How Life Works". At that time (February 2024), I promised to write a review but that never happened for two reasons. First, there were just too many misrepresentations and misinterpretations to cover and I got bored and frustrated trying to document them. Second, I found it very difficult to come to grips with Ball's main thesis, which seems to have something to do with agency or some higher level of control of human development.
I recently mentioned Philip Ball in a post covering a review of The Selfish Gene [The Selfish Gene: a critic and a defender are both wrong!]. This prompted an exchange with Ball in the comments section where the question of how to define a gene came up. That led to a separate post where we discussed that issue: Defining a gene: the selfish Mendelian gene and the molecular gene.
The issue here is whether the product of a gene has to have a well-documented function or whether the mere fact that a stretch of DNA is transcribed qualifies it as a gene. In the latter case, there are thousands of non-coding genes but that number drops to a much smaller number (<1000?) if you require that the RNA product must have a known function supported by direct evidence.
Ball's position is that I shouldn't be so picky about how to define a gene when writing for a general audience. In the comments he claims that he is merely reporting the "standard narrative."
The significant shift in the way we think about RNA in recent decades is reflected in the fact that there are meetings like this one: Cell Press Symposia: Functional RNAs. This is simply the standard narrative today. Larry is of course free to challenge it, but the position described in my book reflects the prevailing view.
Another commenter (SPARC) said, "IMO we should have learned from the ENCODE disaster that narratives do not necessarily reflect the facts coherently and the prevailing view is not necessarily the full truth" and Philip Ball responded with ...
The idea that I'm in thrall to the "standard narrative" of molecular biology and genetics is very funny, though the humour might be lost on this site.
So let's examine whether the "standard narrative" is the view of knowledgeable experts or just the majority view of people who may not have studied the problem in sufficient detail. Let's examine whether Philip Ball is in thrall to the "standard narrative" by looking at the presentation he gave to The Royal Institution in April 2024.
You can skip the first 8 minutes.
I can't possibly go into all the ways in which Phil Ball has been mesmerized by the postgenomics "standard narrative" so let's just pick a few examples.
Central Dogma
The (incorrect) narrative is that the central dogma is DNA→RNA→protein and Ball falls for that narrative hook, line, and sinker (8:00 minutes). The standard postgenomics narrative is that because of the central dogma, the old fuddy-duddies thought that the only role of DNA was to encode proteins. Check that box as well.
(The image on the right is from his talk at 11:12. He introduces it by saying, "Well, here's how that will be expressed within the context of Crick's central dogma.")
Lock-and-key model
The old-fashioned narrative about enzymes is the lock-and-key model where the enzyme and its substrate have rigidly fixed structures that fit together like a lock and a key. Ball repeats this old-fashioned view at 11 minutes and tries to make a big deal about it when he brings up intrinsically disordered regions of a protein later on in his talk.
The modern view of enzyme action is the "induced fit" model where the protein conforms to the shape of the substrate and this is especially true of disordered regions that acquire specific order when they bind to their substrate.1 (The lac repressor is an excellent example. The free protein has a disordered structure at the DNA binding site but it becomes highly ordered when it binds to DNA.) The induced fit model has been in all the best biochemistry textbooks for decades but the standard narrative among the non-experts is the lock-and-key model.
[Note: I assumed that the reason for putting such an emphasis on rigid protein structures was that Philip Ball was unaware of the more dynamic nature of protein structure and that he had fallen for the "standard narrative" that's common in the popular press and on the internet. But I was completely wrong about that as Ball pointed out in the comments. I apologize to Philip Ball for misrepresenting his view.
Ball has written a lot about induced fit models and the fact that proteins can conform their shape to the substrate (or the transition state). In fact, I found a News & Views article he wrote in 1994 (Ball, 1994). He even indicates in that article that this is a "new perspective" on the old lock-and-key model.
Thus, he has known for at least 32 years that the old lock-and-key model is inadequate as an explanation for enzyme activity. There must be another reason why in his talk he chose to emphasize the rigidity of protein structures.]
The postgenomics revolution
The standard narrative is that there have been huge changes in our understanding of human biochemistry and molecular biology since the completion of the human genome project. This is the narrative pushed by science writer Elizabeth Pennisi and philosopher Evelyn Fox Keller. Philip Ball fell down this rabbit hole (11:40). (We've learned some things since 2001 but none of them are paradigm-shifting.)
Number of genes in the human genome
Back in the late 1960s, knowledgeable scientists estimated that the human genome had about 30,000 genes and these same knowledgeable scientists continued to stick to roughly that number right up to the completion of the human genome project in 2001. The initial sequence results predicted about 30,000 genes so the experts were correct. [How many protein-coding genes in the human genome? (2)] [False History and the Number of Genes 2010] [Facts and Myths Concerning the Historical Estimates of the Number of Genes in the Human Genome]
In 1990, at the start of the human genome project, Walter Gilbert made a back-of-the-envelope calculation of the number of genes. He assumed that the average gene was 30,000 bp long and the genome was chock full of genes (no junk DNA). This meant that there were about 100,000 genes. Every knowledgeable scientist knew that this was a ridiculous number but it got inserted into the U.S. Human Genome Report.
The standard narrative is that the experts expected 100,000 genes and they were "shocked" when only 30,000 genes were discovered. Philip Ball repeats the standard narrative (12:20) in spite of the fact that many papers have shown that Gilbert's calculation was wrong and conflicted with what was known back in 1990.
Thousands of non-coding genes
We've known for 50 years that the human genome is pervasively transcribed and the number of different transcripts has risen exponentially in the last few decades.
Knowledgeable scientists have established that most of these transcripts are probably junk RNA that has no biologically relevant function. However, some scientists have pushed the narrative that most of these transcripts must have a function and they don't hesitate to refer to the templates as "genes" in spite of the fact that there's no logical defense of such a definition of a gene.
This idea of more non-coding genes than protein-coding genes has become the standard narrative ever since the ENCODE publicity disaster and in this lecture Philip Ball comes across as an uncritical believer. (The figure is from his lecture.)
The Deflated Ego Problem
Lots of people were upset to find that humans didn't have a lot more genes than a nematode or other mammals and even fewer than some plants, like a banana. I call this phenomenon The Deflated Ego Problem for obvious reasons.
People suffering from deflated egos have a standard narrative where they point out how silly it is to think that complex creatures like humans have fewer genes than nematodes. They play on the fact that a general audience might think this is clearly wrong and sometimes they even do silly things like hold up a banana and ask who is more complex, humans or bananas (12 mins).
The truth is more complicated; we've known for decades that the real issue is not how many genes you have but how they are regulated. It's much harder to explain the truth to a general audience and much easier to recite the standard narrative and take advantage of their ignorance. It will get you a laugh and make you sound intelligent.
(@13:08) And these numbers are often paraded now as a comic example of how wrong expert opinion can be. But I think the really important question is why they were wrong. Did we perhaps have the wrong idea about what role genes were playing?
Alternative splicing
Biologically relevant alternative splicing has been known for decades and it has been well-described in the textbooks for more than forty years. The number of well-defined examples has increased over the years so that we now know that there are at least several dozen examples in humans.
Lots and lots of splice variants have been identified in genomic studies of total RNA. All the experts who have looked critically at this data have concluded that the vast majority of these splice variants are produced by errors in splicing. There are some genuine examples of biologically relevant alternative splicing but those are quite rare.
This is not the standard narrative in science articles for the general public. Instead, the standard narrative picks up on those scientists who claim that almost every gene exhibits alternative splicing. Philip Ball loves most "standard narratives" in the popular press so in his lecture he reports the shocking news that each protein-coding gene can make about six different proteins (16:30). [see: Philip Ball's view of alternative splicing]
1. Here's how I describe it on page 180 of the 2011 edition of my biochemistry textbook. (Similar wording appears in the 2002 edition.)
Hexokinase, citrate synthase, and triose phosphate isomerase are extreme examples of induced fit mechanisms. Recent advances in the study of enzyme structures reveal that almost all enzymes undergo some conformational change when substrate binds. The simple concept of a rigid lock and a rigid key is being replaced by a more dynamic interaction where both the "lock" (enzyme) and the "key" (substrate) adjust to each other to form a perfect match.Ball, P. (1994) The shapes of things to come. Nature 371: 202-203. [doi: 10.1038/371202a0]
I should perhaps be flattered that Larry risks boring his readers by returning once again to his slightly weird obsession with trying to prove that I am an ignoramus about biology. I guess something I said must have unsettled him. But any temptation to feel flattered by the attention is tempered by the fact that he seems to put so little effort into it, or so little faith in readers’ ability to check for themselves. I won’t compound readers’ tedium (or my own) by going through all of these comments, but will trust to some representative responses to make the general point.
ReplyDelete“The (incorrect) narrative is that the central dogma is DNA→RNA→protein and Ball falls for that narrative hook, line, and sinker”. If anyone bothered to look at the video at that point, they’ll see that I use Crick’s own diagram of the CD – literally the one in Crick’s paper – which is more complex than this incorrect one propagated by Watson. I’m not outraged to this sort of thing because I’m inured to it, but it seems a bit disrespectful to his readers (“Oh, they won’t bother to check”).
“The standard postgenomics narrative is that because of the central dogma, the old fuddy-duddies thought that the only role of DNA was to encode proteins.” Well, we’ve been through this endlessly previously, the issue here being about how much we knew at various stages about the extent of noncoding RNA. I think I’ve said all I need say there.
“The old-fashioned narrative about enzymes is the lock-and-key model where the enzyme and its substrate have rigidly fixed structures that fit together like a lock and a key. Ball repeats this old-fashioned view at 11 minutes and tries to make a big deal about it when he brings up intrinsically disordered regions of a protein later on in his talk. The modern view of enzyme action is the "induced fit" model where the protein conforms to the shape of the substrate and this is especially true of disordered regions that acquire specific order when they bind to their substrate.”
Of course I know about the induced-fit picture, and have written about it plenty of times – here for example:
https://www.chemistryworld.com/opinion/snapshots-of-lifes-dancers/3008393.article
Obviously the key issue in the present context is the notion of shape/chemical complementarity. I could have said that the more modern view is that the receptor has some wriggle room to fit the ligand, but that wouldn’t have added anything substantial for a general audience. If Larry thinks that induced-fit takes care of what we now know about IDPs, he is simply not abreast of that field. And frankly, it’s pretty clear that he is not. Some IDPs do fully fold on binding (which is already different from an induced-fit picture). Some do that just a little. Some don’t fold at all. And the key point here – the one that I make – is that this conformational looseness can create a strong degree of binding promiscuity. We could argue then about the extent to which disorder-induced promiscuity was recognized at different points in time, but people who work on IDPs (who I have spoken with a great deal) are pretty agreed in the view that this has emerged as a strong theme in recent decades and helps to make sense of, for example, the behaviour of many eukaryotic transcription factors. But maybe that entire field is also wrong/deluded/ignorant too because once again we knew it all in the 1970s/80s?
And so on. Larry, this game of Gotcha – “You said X but what about Y? Proves you don’t know what you’re talking about!” is very silly, not least because, as you’ll know, there are rather few statements in biology for which one can’t say “what about Y?” To the response that I’m talking to a general audience and so not appending endless caveats and footnotes, you boast that *you* don’t cut corners but just insist on “accuracy, accuracy, accuracy”, and yet we’ve seen many times that this is simply untrue. You misrepresent or just make stuff up, you are sometimes plain wrong, and you state as fact what you know is just opinion. But by all means carry on with your peculiar crusade.
That’s a bit off, Philip. You presented Crick’s picture but then constructed your own misrepresentation around it. Crick was presenting something very specific in that diagram – the possible pathways of SEQUENCE information transfer amongst the main information-carrying macromolecules (DNA, RNA and protein) as far as he considered the evidence supported in 1956/1958. His point was that sequence information could be transferred into protein but couldn’t be transferred back from protein to RNA or DNA. That’s it. The diagram was NOT meant to be a representation of what was known about genes and proteins and how gene expression leads to phenotype and so on. Those are your misrepresentations that you have chosen to dump onto Crick’s point (and his diagram).
ReplyDeleteWe know exactly what Crick meant by the Central Dogma because we can read his words. How sad that you don’t consider Crick’s words should be allowed to speak for themselves. Here’s what Crick said:
On Protein Synthesis p 138 Symp soc exp biol, 1958
The Central Dogma
This states that once 'information' has passed into protein it cannot get out again. In more detail, the transfer of information from nucleic acid to nucleic acid, or from nucleic acid to protein may be possible, but transfer from protein to protein, or from protein to nucleic acid is impossible. Information means here the precise determination of sequence, either of bases in the nucleic acid or of amino acid residues in the protein. This is by no means universally held-Sir Macfarlane Burnet, for example, does not subscribe to it-but many workers now think along these lines. As far as I know it has not been explicitly stated before.
I stand by my words. Which is to say, I present the CD and then I talk about how it has been construed in a wider picture of how phenotype is connected to genotype. I explain this in some detail on pages 182-3 of my book. It is perfectly obvious that one is not going to get that level of detailed exposition in a 1-hour talk. In my book I explain in what sense the CD still holds today.
Delete@Philip Ball: My intention wasn't flattery but respect. I respect the fact that you are willing to engage in this debate.
ReplyDeleteMany people are making the same errors of interpretation that you make but most of them aren't courageous enough to try and defend them.
With respect to to the central dogma, you introduced the idea and said that it is wrong. Then you described transcription and translation and the activity of enzymes. For the benefit of readers, I've added the figure that you presented after this and introduced with "Well, here's how that will be expressed within the context of Crick's central dogma."
Readers can judge for themselves whether this is an accurate representation of Crick's central dogma of molecular biology.
I'm glad to hear that you respect that Larry. And I want to try to engage in the debate constructively and civilly. Sometimes that feels hard, as for example when, as here, you so clearly misrepresent me. I do not say Crick's CD is wrong. In fact I say that plenty of people have devoted considerable time and energy to trying to prove that it was wrong, and then imply, I believe, that this is time and energy not really well spent. In my book I expand on this, as you know, explaining how specifically the CD remains fit for purpose today but why also we should not let it seduce us into believing a more general narrative about information flow.
DeletePhilip Ball, cutting to the chase what percentage of the genome do you think is functional?
ReplyDeleteReally Neil? After I've just been talking about Gotcha comments? ;-) Because you see, I don't think this is cutting to the chase at all - as if it all comes down to this. I don't think anyone knows what percentage of the genome is functional. And while I recognize that some consider this an interesting or important thing to specify, I don't. I think there is a significant proportion of the genome that is of no use to the cell. I think that more of it *is* of use than we thought some decades ago, and I'd not be at all surprised if that proportion continues to increase somewhat. I also think that there is now good reason to doubt that there is a clear distinction between what is functional and what is not, or a clear way of defining "functional". And for me, that is where the interesting questions lie: in elucidating how we think about functionality, and how function is realised. I'm interested in how it all works.
Delete@Philip Ball re: "I stand by my words".
ReplyDeleteI don't think that's acceptable. You didn't present the Central Dogma at all - you showed Crick's diagram associated with the Central Dogma. This obviously doesn't represent the meaning of the CD without an explanation of what the arrows signify. You chose not to do this (and so you didn't "present the CD") and instead used the diagram as backdrop to your misrepresentation; one that is quite insulting to Crick in fact.
Much as you fault James Watson on p 183 of your book, perhaps for you also "rigor matters less than rhetoric".
Why is it insulting to Crick given that I don't say the CD is wrong? Where is this insult? What I am saying is that we should not be encouraged to believe - as I think some like Watson would have us believe - that the CD implies something more general about biological information flow. I chose my words carefully, but of course I cannot chose my readers/listeners.
ReplyDeleteIt's insulting because it's a complete misrepresentation of Crick's Central Dogma, and since you use his diagram as a backdrop to your rhetoric about how all those silly scientists thought that genes produced proteins that produced the phenotype and so on the impression is that this is all part of Crick's misunderstanding.
DeleteIt's very easy to state Crick's Central Dogma concisely, something like: "biological sequence information can flow from polynucleic acid into protein but can't flow out of proteins in the other direction". The Central Dogma has got nothing to do with what genes do, or the relationship between genotype and phenotype and all the other stuff you tacked on while referring to Cricks diagram (with your inappropriate add-ons).
No, because I never said "And Crick also said that the CD meant this...". I can do nothing about how your preconceptions have distorted your interpretation. My experience has told me that academics sometimes get confused when the framework is shifted, but that's not my problem.
ReplyDeleteI mean, it sort of is my problem because it means I have to deal with stuff like this, but I don't imagine I can do much about it.
DeleteOK, let me try to be more charitable. The point is this. "Here is Crick's CD. In the common view of the genetic blueprint, the corollary is that ultimately the information therein dictates the phenotype [and here's what that would look like if we add it into this picture]. But in fact this is not what the information flow in organisms looks like." I suspect that this might be more clear to someone coming at this afresh, than it is to someone coming at it expecting to find another misrepresentation of the CD. In my book I am somewhat more explicit with that second possibility in mind. In my talk I don't have time to deal with misconceptions that are unlikely to arise in the lay person in the first place.
ReplyDeleteI am not sure that Crick's CD is the best context in which to present a discussion of the expression and function of non-coding RNAs.
ReplyDeleteI'm quite keen on digging into the way in which narratives and framings arise, in the manner of Lily Kay's Who Wrote The Book of Life?, so this may be where we differ.
Delete@Philip Ball: I'm quite keen on digging into the ways in which narratives and framing have influenced the work of Philip Ball.
ReplyDeleteIt would be nice to learn how he fell for the false narratives about alternative splicing and predictions about the number of genes. I'd love to know why he picked up on the non-coding narrative that conflicts with any reasonable definition of a gene. And why, after decades of evo-devo, did he find himself susceptible to the deflated ago problem?
All we've heard so far is a weak attempt to defend his misrepresentation of the central dogma and the fact that he knew about induced fit and the roles of disordered structures but didn't think it was necessary to mention that to a general audience.
Monday, November 09, 2026 11:51:00 AM
Ah. What we've heard is that, as anyone can easily go and check in my talk, I do not say the CD is "wrong", as you claimed. Simple matter of fact. What we've heard is that anyone who thinks the induced fit model explains how many IDPs bind hasn't kept up with the literature. Simple matter of fact. But you address neither of those points. Odd.
ReplyDeleteThe rest is noise.
@Philip Ball says, "I do not say the CD is 'wrong', as you claimed."
ReplyDeleteCrick's Central Dogma of molecular biology is as correct today as it was in 1958 and 1970. Is that what you meant to say?
At around 17:40 you say, "In other words the information flow here isn't, as the central dogma at least implied, isn't all from the bottom up - from DNA to RNA to proteins." Did you forget to say "but the central dogma is not wrong, it is still right"?
At 16:00 you say, "It's actually even worse than this, Crick's central dogma said that genes encode proteins, and by that we mean that genes actually program the proteins with particular shapes so that they can go and do their specific jobs." You then go on to say that each gene can encode multiple proteins. Did you forget to remind your audience that this misrepresentation of the central dogma is still not wrong - it is right?
I wasn't claiming that the induced fit model explained every example of intrinsically disordered regions. My colleagues in the Department of Biochemistry at the University of Toronto have been working on intrinsically disordered regions for more than two decades so I am reasonably familiar with the literature.
I'm attaching a review written by my friend and colleague Julie Forman-Kay.
The point I was making is that you exploited an old-fashioned lock-and-key view in order to set up your audience for the revolutionary idea that some proteins have intrinsically disordered regions. I assume that you brought up the old view of lock-and-key because you were unaware of the more modern views about dynamic protein shapes and induced fit.
In other words, it was another example of you falling for the "standard narrative" that's expressed in the popular press and on the internet.
You could have made a point about intrinsically disordered regions without setting up a strawman about rigid structures.
Forman-Kay, J.D. and Mittag, T. (2013) From Sequence and Forces to Structure, Function, and Evolution of Intrinsically Disordered Proteins. Structure 21(9): 1492-1499. [doi: 10.1016/j.str.2013.08.001]
Ah, I see! So when you said that I had said the central dogma is wrong, what you meant was that I failed to say that it is (still) right! Let me explain the narrative: I point out that some folks get hung up on arguing whether the CD is right or wrong, whereas the point I make is that, for the purposes of thinking about the informational (and causal) flows in cells and organisms more generally, that's not really very important. I think it is clear that your conviction that this was going to be yet another "The central dogma is wrong" story has prevented you from being able to follow what was said. I see this happen a lot in science.
ReplyDeleteJulie's review is great, and I've had the pleasure of discussing IDPs with her previously. It does a great job of showing how, although some notions of disorder have been discussed since the mid 1980s, things have really shifted in the past 20 years or so. As she says, "Intrinsically disordered proteins (IDPs)...are a challenge to structural biology due to... their deviation from the dominant structure/function paradigm... Intrinsically disordered proteins represent a significant challenge to the previously described body of structure/function correlations... the paradigm of structure/function correlations, while facilitating significant biological insight, slowed the acceptance of the biological role of highly dynamic and disordered protein states... Based on these and many other important contributions over the past 20 years, a set of characteristics of intrinsically disordered proteins and regions (IDPs/IDRs) has been established that provide a context for functional correlations to augment those described for folded protein structures." I did not mention the induced-fit view because it is not really relevant to this shift in how we think about folded vs disordered structures - it was primarily a modification of the rigid lock-and-key view of folded proteins (and of course it was, given that Koshland presented it in 1958), and as such, was a part of the "dominant structure/function paradigm" that Julie talks about. [If you disagree with that, I assume you're going to castigate Julie for talking about a "dominant paradigm" that in fact has been anything but since at least 1958?] As such, it would have added nothing substantial to the story about the emergence, in the past several decades, of the story about IDPs, as Julie has so eloquently described.
I think you know all this too, deep down, but you thought you'd scored an easy point and now you've tied yourself in knots.
Why, by the way, *assume* that because I don't mention the induced-fit model, I must not know at all about something so basic? Why not wonder if (as I've now explained) I simply had no reason to do so? It only takes a quick Google search to answer that question. Why this insistent lack of generosity in your assumptions, whereby science writers are just total ignoramuses? (Let alone ones who have written highly cited review articles about protein hydration and dynamics, but hey.)
ReplyDelete@Philip Ball: Let's see if you can give a straightforward answer to a simple question about the central dogma.
ReplyDeleteMy position is that the central dogma is as correct today as it was in 1958 and 1970.
What is your position? Is it still right today just as it was back then?
Oh yeah, I'm happy with it, despite efforts to "prove it wrong". I even challenged "CD is wrong" narratives after that recent work on an enzyme that makes DNA using its own structure as a template [https://www.science.org/content/article/scientists-stunned-fundamentally-new-way-life-produces-dna]. But you knew that, surely, as I said so in my book?
Delete@Philp Ball: With respect to your knowledge about "induced fit" I apologize sincerely for misrepresenting your view. I did do a Google search and I found a News and Views article that you wrote in 1994 about induced fit.
ReplyDeleteYou described how proteins can adapt their structures to their substrates and you said, "This sensitivity of host structure to guest binding has been a theme of supramolecular chemistry since the 1960s ..."
You also said, "The second new perspective on the lock-and-key principle comes from the fact that enzymes do not simply bind - they perform catalysis. Good substrate binding alone does not guarantee this function; indeed, if the effect were simply to stabilize the substrate, it's reactivity would be lowered. The true influence of the protein is rather to stabilize the transition state (TS) of the reaction relative to reactants or products."
So, it looks like I was completely wrong about whether you knew about structural dynamics and induced fit models. You've known about them for 32 years. There's probably another reason for why you put such an emphasis on rigid protein structure in your talk.
I've edited my post to reflect my error. I apologize again.
Thank you. My emphasis was on "structure", not "rigid", as per Julie's review article. I don't believe (unless I've forgotten) that I mention rigidity at all, but just folding and structure.
ReplyDelete