Correcting a Quote Mine on Deleterious Mutations

Conversation
1

“In summary, the vast majority of mutations are deleterious. This is one of the most well-established principles of evolutionary genetics, supported by both molecular and quantitative-genetic data.”
Keightley P.D. and Lynch, M., Toward a realistic model of mutations affecting fitness, Evolution 57 (3):683–5, 2003

I’ve personally replied to this misquote several times now, here it is again:

The 2002 Keightley and Lynch paper, entitled “TOWARD A REALISTIC MODEL OF MUTATIONS AFFECTING FITNESS,” is a response paper to a mutational accumulation experiment done by Shaw et al.–this is that whole ‘peer review’ process going on. The “other” scientists claimed that their MA experiment yielded 50% ADVANTAGEOUS mutations–which every model of evolution denies is possible, including Neutral theory. MA experiments artificially prevent natural selection from occurring by controlling mating, population size, and providing unlimited food/resources. The entire paper is referring to mutations in coding regions as is the Shaw et al. experiment. Quotes from the paper that you ignored:

“However, in all taxa examined so far, average values of C are in excess of 0.7 (e.g., Ohta 1995; Eyre-Walker et al. 2002), implying that the majority of amino-acid altering mutations are deleterious.”

“There is nothing obviously unusual with respect to A. thaliana in this regard. Wright et al. (2002) and S. Wright (pers. comm.) have recently investigated constraint in the protein-coding genes of two species of Arabi- dopsis, A. lyrata (an outcrosser) and A. thaliana (a natural inbreeder), using an outgroup to infer lineage-specific constraint. Estimates for C are 0.88 in both species, despite their different systems of mating; C is likely to underestimate the fraction of amino-acid mutations that are deleterious due to fixation of advantageous amino-acid mutations and purifying selection acting at synonymous sites (Eyre-Walker et al. 2002).”

To follow, quotes from the same author:

Of note, GE proponents selectively misquote these works and apply the authors’ quotes to the entire genome when only the coding-regions are specifically addressed. For example:

The GE proponent quotes Eyre-Walker, A. & Keightley (2007):

“The first point to make is one of definition; it seems unlikely that any mutation is truly neutral in the sense that it has no effect on fitness. All mutations must have some effect, even if that effect is vanishingly small.”

The full quote in context (ibid.):

“The first point to make is one of definition; it seems unlikely that any mutation is truly neutral in the sense that it has no effect on fitness. All mutations must have some effect, even if that effect is vanishingly small. However, there is a class of mutations that we can term effectively neutral . These are mutations for which Nes is much less than 1, the fate of which is largely determined by random genetic drift. As such, the definition of neutrality is operational rather than functional ; it depends on whether natural selection is effective on the mutation in the population or the genomic context in which it segregates , not solely on the effect of the mutation on fitness.”

These definitions from Eyre-Walker, A. & Keightley (2007) are specifically referencing mutation accumulation (MA) assays which historically interrogated only coding-region mutations. More recent MA experiments often characterize whole genome mutations such as in Dillon, M. M. & Cooper, V. S. (2016). Eyre-Walker, A. & Keightley (2007) go on to say:

“Unfortunately, accurate measurement of the effects of single mutations is possible only when they have fairly large effects on fitness (say >1%; that is, a mutation that increases or decreases viability or fertility by more than 1%)”

“In hominids, which seem to have effective population sizes in the range of 10,000 to 30,000 (Ref. [29]), the ratio dn/ds is less than 0.3 (refs [29],[42]), and this suggests that fewer than 30% of amino-acid-changing mutations are effectively neutral. ”

“The proportion of mutations that behave as effectively neutral occurring outside protein-coding sequences is much less clear.”

“In mammals, the proportion of the genome that is subject to natural selection is much lower, around 5% (Refs [55]–[57]). It therefore seems likely that as much as 95% and as little as 50% of mutations in non-coding DNA are effectively neutral; therefore, correspondingly, as little as 5% and as much as 50% of mutations are deleterious. ”

12 Likes
2

“Unfortunately, accurate measurement of the effects of single mutations is possible only when they have fairly large effects on fitness (say >1%”

“In hominids, which seem to have effective population sizes in the range of 10,000 to 30,000"

Nearly neutral is 1/(2Ne), which for humans is between 1.7x10^-5 and 5x10^-5. That means we can only accurately measure fitness effects 200-600x neutrality.

2 Likes
3

So? They still said what they said.

No, that’s not what the authors said. They said that the vast majority of mutations (no qualifier) are deleterious.

Maybe you feel they shouldn’t have used the wording they did, but that doesn’t make it a quotemine just because you don’t like what they said.

Once again, none of this quoted material acts to render what I quoted inaccurate, or in any sense a quotemine.

That is not what the authors said. You still haven’t shown any quotemining here.

Your quote of this section is puzzling, since it agrees exactly with what I (and Sanford, really) have been arguing all along. It is also exactly what Kimura wrote in his 1979 paper.

Sure, but that doesn’t mean we cannot learn anything about the general DFE for all mutations (e.neutrals included!). This is addressed in our joint response article.

Yes, but effectively neutral is not strictly neutral. They still have an impact.

In this context, they are using the word “deleterious” to mean “not effectively neutral”. That’s very clear in the sentence you just quoted. But as we already know, effectively neutral does not mean “no fitness impact at all”, as you seemingly think it does. Based upon Sanford’s analysis, we still know that the vast majority of effectively neutral mutations will be deleterious. I suggest you really read his book closely.

4

Nearly neutral is 1/(2Ne), which for humans is between 1.7x10^-5 and 5x10^-5. That means we can only accurately measure fitness effects 200-600x neutrality.

Sure and if no observed difference is found in the fitness of the organism, then one cannot claim the effects are deleterious–which is what GE attempts to do.

5

Exactly. You measure 0.00001 deleterious, it may well be 0.00005 beneficial. You can’t say. All of GE rests on an assumption they are mechanistically prevented from validating.

1 Like
6

@PDPrice Maybe it’s because I haven’t finished reading the book, or don’t understand mutations or just the semantics here, but I don’t understand how these claims undermine GE instead of strengthen them.

How do population geneticists validate it if they’re saying you’re unable to?

7

GE is specifically contingent upon the presence of overwhelmingly deleterious mutations–which it cannot show exist. GE proponents take quotes from studies looking at amino-acid changes and claim this deleterious propensity applies to all parts of the genome. However, the available data do not suggest this is the case and neither do experimental mutational accumulation assays.

When we do look at non-coding parts of the genome, we find that only portions of it are under selection. When we look at mutations in the non-coding parts of genome, they overwhelming confer no observable fitness difference to the organism (especially in humans). One might wonder, given the data, how it’s possible to justify the claim that “mutations are heavily skewed toward being deleterious” when it’s not observed.

2 Likes
8

You were asked by @glipsnort to present evidence that most mutations of very small effect are deleterious:

I don’t know what you’re trying to say here. If you have evidence that most mutations of very small effect are deleterious, present it. You’ll be the first to do so.

You responded with this:

All the data we have on mutations strongly and clearly show a massive skew toward deleterious. That’s why:

“In summary, the vast majority of mutations are deleterious. This is one of the most well-established principles of evolutionary genetics, supported by both molecular and quantitative-genetic data.”
Keightley P.D. and Lynch, M., Toward a realistic model of mutations affecting fitness, Evolution 57 (3):683–5, 2003

It’s contextually quote mined because the study is in reference to only coding-region mutations–which do not represent most mutations. Because the coding-regions represent only a small fraction of the total genome, this misrepresents the total mutational landscape. I included subsequent quotes from the same authors in a different paper which correctly contextualizes the quote:

“In hominids, which seem to have effective population sizes in the range of 10,000 to 30,000 (Ref. [29]), the ratio dn/ds is less than 0.3 (refs [29],[42]), and this suggests that fewer than 30% of amino-acid-changing mutations are effectively neutral. ”

“The proportion of mutations that behave as effectively neutral occurring outside protein-coding sequences is much less clear.”

“In mammals, the proportion of the genome that is subject to natural selection is much lower, around 5% (Refs 55–57). It therefore seems likely that as much as 95% and as little as 50% of mutations in non-coding DNA are effectively neutral; therefore, correspondingly, as little as 5% and as much as 50% of mutations are deleterious. ”

Eyre-Walker, A. & Keightley, P. D. The distribution of fitness effects of new mutations. Nature Reviews Genetics 8 , 610–618 (2007).

It’s quite clear that the DFE for coding-regions is vastly different than the DFE for non-coding regions. So, while it is appropriate to claim “70% of amino-acid changing mutations are deleterious,” it is factually incorrect and contextually dishonest to say that “Most mutations are deleterious.” This is because 1-2% of the human genome is coding and 98-99% is non-coding–meaning that your postulate of

“All the data we have on mutations strongly and clearly show a massive skew toward deleterious.”

Actually means 70% of 1-2% of all the data agree with your statement while 50-95% of 98-99% of the data disagree with your statement.

I have highlighted this issue on more than one occasion and I’m hoping that you decide to stop using this as a talking point in the future.

11 Likes
9

GE relies on the assumption that nearly neutral mutations are overwhelmingly deleterious. I mean that literally, in that the nearly neutral deleterious ones overwhelm all beneficial mutations (both nearly neutral and selectable). But for humans, the limit of detection is at least 200x the effect size of effective neutrality, so it is impossible to demonstrate that this is actually the case, they simply assert it is without direct evidence.

Of course, there is also the problem that at the upper end of possible neutrality limits for humans, a 10% reduction in fitness would take over 52000 years, and that is with no beneficial mutations at all and one ‘just barely’ unselectable mutation per generation. Any plausible model would have the average fitness effect per generation much closer to zero than that, though, since a) there must be some beneficial neutrals, even if there aren’t b) the average neutral would be much less than the limit. So realistically you’re looking at closer to 1%/50ka, which is well within the plausible frequency of selectable beneficial mutations. So even if we take their argument entirely at face value, it is still basically meaningless.

The short version is that GE is a mathematical argument made by people who are bad at math.

5 Likes
10

That’s right, as far as I can tell.

9 Likes
11

I am taking a crash course on molecular evolution and this was talked about. Glad to see my tutor’s utterances are in line with available evidence.

1 Like
12

Give them a break. Maths is hard :wink:

2 Likes
13

You can simply ignore those statements, since they’re made by people who are actively refusing to understand genetics as an information system, rather than a jumble of arbitrary molecules. These kinds of statements can be categorized as “deliberately missing the forest for the trees”.

1 Like
14

Heh. It’s always interesting when a YEC with no relevant scientific training at all declares we should ignore all the inputs from professional geneticists, many with PhDs, simply because the YEC doesn’t understand what is being discussed. :slightly_smiling_face:

10 Likes
15

Oh and, let’s also recall that there is such a thing as both a local and global optimum for particular proteins with a particular function, so there is some point at which no further beneficial mutations are possible, leaving only neutral and deleterious ones, which directly implies the physical impossibility of a constant DFE for coding regions, confirmed by something like half half a century of work in directed evolution of enzymes.

6 Likes
16

Totally agree. I just think it’s fascinating that we can allow ridiculous concessions like immutable DFE and Genetic Entropy is still not supported. It doesn’t even work in mutational accumulation experiments with heavily attenuated natural selection, hypermutable bacteria strains, and ridiculous coding to non-coding ratios over thousands of generations.

4 Likes
17

‘Ignore people with math, they’re mean!’ Great rebuttal, real quality work. Seriously man, you can’t make a mathematical argument, which GE is, then not provide any math to support it and ignore all the math that refutes it. That’s not how this works.

And @thoughtful, notice that Paul’s response is an instruction for you to stick your fingers in your ears? This is not how people with understanding respond. But if it makes you feel any better, at least you aren’t the only one that doesn’t understand the objections, because I’d say you understand it as well as Paul does.

7 Likes
18

I didn’t read it exactly that way, but I’m not sticking my fingers in my ears. I’m taking it all in. I definitely don’t understand everything - probably not most of it. But I can ask questions. And this is an interesting debate. I like it when people actually provide substantive responses so I can learn something.

Looking at it the other way, is it possible to demonstrate that nearly neutrals are beneficial instead? I still don’t understand how this works.

If someone would like to answer why population geneticists say we’re degenerating then if there’s no direct evidence, I’d appreciate it.

3 Likes
19

She understands it much better, because she often understands that she doesn’t understand what she doesn’t understand. That’s much better than thinking one understands things one does not.

5 Likes
20

Thank you. But sometimes I do think I understand much more than I do. If I really overstep what I understand, you guys will let me know. :sweat_smile: I’m trying to do better so it doesn’t happen as often.

Bacteria are a different ballgame.