See the three papers linked above.
Let’s talk the first. Is there something unclear about the statement that “SHM can also aberrantly target proto-oncogenes and cause genome instability,” or the data supporting that conclusion?
I think this misses the point @mercer. Let us state the obvious: there is a disagreement between you and I about this. One of us is correct, and one of us is wrong, or both of us are incorrect. Either way, none of this presents a challenge to the notion that mutations are “random”. Whoever is correct on the precise ratio of beneficial to negative mutations here, the mutations are still random. Right? If so, it is really just a rabbit trail for the larger conversation.
Also, do you really want to set up the conversation so that if I was right it would be challenge to evolutionary theory? Even if, in this particular case, I’m wrong, are you sure this is true in all cases? I’m certain our definitions of “random” are not so fragile, nor is evolutionary science. Basically, whether or not specific mechanisms increase the ratio of beneficial to negative mutations, they are not fully predictable, and in this sense they are “random.”
“Hot spots for hypermutation” covers it just fine. An increase in random mutations in limited regions of the genome is still an example of random mutations, at least in my book. As always, it is helpful to contrast this concept with what we would consider guided mutations (within the biological sciences) which are specific changes to specific bases in direct response to an environmental cue. For example, if all bacteria mutated a specific base that conferred antibiotic resistance in the presence of antibiotic, and only that base and only in response to the presence of the antibiotic, then this would be a non-random mutation.
1 Like
Correct. Having said that, I think it’s a very important point to understand the evolution of such processes through understanding their mechanisms.
Right, except that my position is that the ratio simply cannot be computed. Given the seriousness of the side effects, it could even be a net negative in our first-world environment.
That seems to be the way that you are setting it up, by insisting on the specificity of the process!
Not at all! It would be fascinating if you weren’t wrong in this case or any analogous one.
You seem to misunderstand my motivations, and you strangely keep insisting that I have a calculation to compete with yours. I don’t. I’m saying there’s insufficient information to calculate and that we clearly know that the process is merely selective, not specific. That’s what makes the ratio impossible to calculate.
Metaphorically, it’s a shotgun, not a laser beam. It’s what we would expect evolution to produce.
Mine too.
1 Like
It’s often the case that non-specialists do not use terminology in the same way as specialists. Getting hung up on that point is just an uninteresting debate about semantics. Much more interesting is to translate his request to specialist terminology and then address the actual content.
1 Like
@EricMH if you follow this conversation, @Perry_Marshall decided to call me out and say that I am being equivocal and imprecise on my own definition of “random.” That was his decision. He, a “non-specialist”, was attempting to take a “specialist” to task on my use of technical language in my area of expertise. Now, he has gone silent.
He can use what ever idiosyncratic definition of “random” he wants, but it is not valid to continually complain that other people do not use his definition.
As I understand it, this is merely about terminology, about the use of the word “random.” I can’t see a substantive disagreement that has arisen here, except confusion that arises from misunderstanding of what the term means. Have you seen any substantive disagreement?
1 Like
That reminds me, I still need to look at your book.
But if you write an article about random mutations, it would be extremely helpful to explain exactly what you mean. In the same way, if Marshall chooses to use his definition, he should explain, at least in a footnote that he is aware that he isn’t using the technical definition of random common in computational/mathematical scientific circles.
Leaving your definition of random unexplained will leave a lot of people thinking that you mean each mutation is equally unlikely.
1 Like
I explain it all the time, but I agree that this is a common point of confusion. @Mark, I think we’ve come to understand each other.
1 Like
As someone with a good understanding of the math of random variables but only a rudimentary understanding of the biology, I find this very helpful. Specifically, I read it as focusing on whether or not there is a causal relationship between the environment and the mutation that biases for mutations which increase relative fitness for that specific environmental change.
Whether or not those mutations are modeled as random variables is a red herring* for me (not for OP, I understand). Instead, the key issue is whether the modelling indicates a causal relation between environmental change and mutation which benefits for that change. Experimentally, we can assess causation by manipulating the environment in specific ways and then looking for specific mutations in response, namely those mutations which benefit the organism for that environmental change. @T_aquaticus’s examples of vaccines and antibiotics (in a later post) are the sort of manipulations I have in mind.
As I think of it, for this causal mechanism relating benefit to specific environmental change, it is not enough that the rate of mutation increase in the presence of a stimulus. Nor is it enough that mutations favor those that maintain the viability of the organism, eg by biasing the change to certain base pair changes or to certain regions of the genome. Rather there has to be that specific causal relation of environmental stimulus to (pre-selection) mutation that benefits for that specific stimulus.
ETA: clarifications
*ETA 2: I understand that detecting a causal relation will often involve modeling with random variables. In my post, I use “red herring” to try to indicate that I see this as a diversion from the central issue of communicating what “random mutation” means as a technical term in evolutionary biology.
That’s the approach biologists of the past have taken. The most famous examples are the plate replica experiment run by the Lederberg’s and the fluctuation test run by Luria and Delbruck. In both cases they were able to demonstrate that the mechanisms producing beneficial changes were independent of the environment, at least within the statistical tests used in science.
The beneficial changes cited by the modern Extended Evolutionary Synthesis are also independent of the environment. Going from memory, Shapiro would cite beneficial changes caused by transposon insertion as an example of non-random mutation when in fact they are just as random as any other mutation. Transposons insert in many places in the genome, some of which will be beneficial while others are neutral or detrimental. Shapiro was not able to cite examples where the environment caused a specific insertions instead of insertions that occurred throughout the genome.
I suspect that this discussion might be better served by understanding what biologists would consider non-random mutations. Discussions of randomness tend to drift into philosophical and metaphysical concepts that just aren’t present in the actual science. Biologists are looking at experimental results, and they are not interested in defining or defending a metaphysical view of randomness, at least as their scientific conclusions are concerned.
3 Likes
Yes, he’s clearly talking about what you call ‘uncertainty’. And there is certainly a great difference between his take on evolution and yours, so there is definitely material for a substantive and interesting discussion here, which I’d like to see, so I’ve been nudging the ball forward. But, it doesn’t look like that is going to happen. The thread has been caught in another definition debate. So I’ll leave this thread alone.
2 Likes
I just posted this article, and am linking here as related: