More Recent Comments

Thursday, June 13, 2019

Reactionary fringe meets mutation-biased adaptation: Introduction

This is the first of a series of guest posts by Arlin Stoltzfus on the role of mutation as a dispositional factor in evolution.



Reactionary fringe meets mutation-biased adaptation: Introduction
by Arlin Stoltzfus

Theoreticians often formulate mathematical or computational models with the aim of exploring (or justifying) behavior anticipated from pre-existing verbal theories.  Yet, the resulting formalisms may exhibit behavior that was not expected.  Indeed, sometimes the model breaks the theory.

Reactionary fringe meets mutation-biased adaptation

Introduction
1. The empirical case
2. Some objections addressed
3. The causes and consequences of biases in the introduction process
4. What makes this new?
In the 1970s, deterministic chaos was recognized in a number of dynamic models, including the Lotka-Volterra model used by ecologists for decades to illustrate notions of control via feedbacks between predator and prey abundance.  Depending on the delay in feedback, the system either oscillates predictably, crashes, or becomes chaotic.

Before the chaotic realm was named, characterized, and publicized, surely many researchers stumbled upon it, either when looking at data, or while working out numeric examples.  However, this did not elicit the phrase "Eureka!  I have discovered deterministic chaos!"  Chaotic dynamics did not fit with ideas about "control" or "feedback."  It existed in the world of nature but not in the world of science, even though it involves no new underlying processes.

In evolutionary genetics, the breeder's equation for change in a quantitative trait under selection once justified neo-Darwinism: invoking selection as the creative principle and source of direction in evolution had a rigorous mathematical basis, given abundant infinitesimal variation, and assuming that everything is a continuous trait (see Gould's excellent, well documented analysis in Ch. 4 of Ever Since Darwin, or ca. p. 140 of his 2002 book The Structure of Evolutionary Theory).



What Darwinism means for Ernst Mayr (Mayr and Provine 1980 p. 3).



Yet, after Lande and Arnold (1983) derived the multivariate generalization of quantitative genetics for the simultaneous change in multiple traits, Δz = Gβ, quantitative geneticists began to acknowledge that, in the words of Steppan, Phillips and Houle (2002) "Together with natural selection (the adaptive landscape), [the G matrix] determines the direction and rate of evolution." This new verbal theory, in which the rate and direction of evolution are jointly attributed to selection and standing variation, does not correspond to the old verbal theory, even though quantitative genetics is the branch of mathematical theory most closely aligned with neo-Darwinism, literally assuming abundant infinitesimal variation in every trait.

That is, Darwin's verbal theory led to Fisher's mathematical theory, then further developments along with empirical results (e.g., Schluter, 1996)) led to conflicts with the original verbal theory. As a result, quantitative genetics now tells us something different: the verbal theory of selection as a governing principle or independent shaping force, invoked by generations of neo-Darwinians, is mathematically impossible.

This series of posts relates to another unexpected twist that arises from another archaic ruling-principle theory: the view that the course of evolution largely reflects innate tendencies that shape variation, with selection playing only a minor role, traditionally known as orthogenesis.

A classic argument from Fisher and Haldane (based on their mutation-selection balance equation) says that variation-induced trends cannot be a cause of direction: for mutation to overcome the opposing force of selection would require abnormally high mutation rates. Mutation biases can influence neutral evolution, but otherwise, the only kind of bias that could possibly be influential is an absolute constraint distinguishing possible from impossible variants.

Yet Yampolsky and Stoltzfus (2001) used computer simulations of a simple 2-locus model, along with mathematical formulas based on origin-fixation dynamics, to show how parallelisms and trends may arise from mutational and developmental biases in variation, without requiring neutral evolution, absolute constraints, or high mutation rates.

We will delve into this theory later. For now, the important thing to note is the crucial prediction that the changes involved in adaptation may show the effects of modest quantitative biases in mutations with ordinary rates (e.g., transition-transversion bias in nucleotide substitution mutations).

When this theory was proposed, data on molecular changes involved in adaptation were rare—not sufficient to support statistical hypothesis-testing. In recent years, the data have become much more abundant, and we are seeing the predicted effect in both experimental adaptation (e.g. MacLean, et al. (2010), Couce, et al. (2015), Sackman, et al. (2017)), and more importantly, in retrospective analyses of natural adaptation (e.g., Payne, et al. (2019), Storz, et al. (2019), Liu, et al. (2019), and Stoltzfus and McCandlish (2017)).

That is, we are witnessing the establishment of a fundamental new principle of evolution, a principle that was not just unexpected, but rejected by the architects of the Modern Synthesis.

Thus, it was an odd choice for Trends in Ecology and Evolution (TREE) to publish a deeply deceptive article that attacks this new idea, from a pair of authors so unfamiliar with the topic that they literally mis-define "mutation bias." According to the authors, there are no new principles here, only "standard evolutionary theory" from Fisher, Haldane and Kimura. The appearance of mutation-biased adaptation is illusory, they argue, claiming that the evolutionary biases are due to selection.

What motivated such a gratuitous attack? A colleague who described the paper as "an abomination" assigned it to one of his advisees to study as an example of what a really bad paper looks like. How did it get published? Why didn't the editor get critical reviews?

The answers relate to a dispute between advocates of an "Extended Evolutionary Synthesis" or EES, and defenders of a re-branded version of the Modern Synthesis called SET (Standard Evolutionary Theory). The authors of the hatchet piece are members of a fringe movement dedicated to arguing that (1) SET automagically updates itself to include valid new thinking, and (2) there is no valid new thinking, because anything that seems new actually traces back to important dead people. Thus, these guardians of orthodoxy were obliged to undermine the novelty and importance of mutation-biased adaptation, while at the same time claiming it for SET.

This peculiar set of circumstances—an unorthodox theory, powerful new results, and the backlash from reactionaries—defines a series of posts that Larry has offered to host here on SandWalk. The plan for the series is as follows:
  1. The empirical case. Results from adaptation in the lab, and retrospective analyses of adaptation in nature, show that the changes involved in adaptation are enriched for mutationally-favored changes.
  2. Some objections addressed. The evidence now available rules out the possibility that the observed evolutionary biases are due to a cryptic bias in fitness that happens to align with the mutational bias.
  3. The causes and consequences of biases in the introduction process. The theory of Yampolsky and Stoltzfus (2001) addresses mutation biases, developmental biases, and effects of connectivity of genetic networks (invoked in the self-organization literature)
  4. What makes this theory new. For a very long time, mutation-biased adaptation was not anticipated, due to the "gene pool" assumption that evolution begins with standing variation.
  5. A diversion into the EES-SET culture war. Issues relevant to navigating high-level disputes in evolutionary biology are addressed in a series of 4 posts.
    5.1. Thinking about theories.
    5.2. The Modern Synthesis (1959 - 1969).
    5.3. The abuse of history.
    5.4. Synthesis sophistry.
  6. What "limits" adaptation? What makes adaptation something other than an ideal process in which the best genotype arises in infinite time after all possibilities have been tested?
  7. Future directions. Abundant opportunities exist to build a broader empirical and theoretical understanding of the evolutionary role of biases in the introduction of variation, and to leverage this role in evolutionary inference.
The posts will be released every few days, and will be linked in to this Introduction page (so you can bookmark this).

So, please join in the discussion, and invite your colleagues to do the same.

Next post: Reactionary fringe meets mutation-biased adaptation. 1. The empirical case.


Couce A., Rodríguez-Rojas A., and Blázquez J. (2015) Bypass of genetic constraints during mutator evolution to antibiotic resistance. Proc. Biol. Sci. Apr 7;282(1804):20142698 [doi: 10.1098/rspb.2014.2698]

Lande, R., and Arnold, S.J. (1983) The measurement of selection on correlated characters. Evolution, 37:1210-1226. [doi: 10.1111/j.1558-5646.1983.tb00236.x]

Liu, C., Leighow, S., Inam, H., Zhao, B., and Pritchard, J.R. (2019) Exploiting the 'survival of the likeliest' to enable evolution-guided drug design. bioRxiv 557645; [doi: 10.1101/557645]

MacLean R.C., Perron G.G., and Gardner A. (2010) Diminishing returns from beneficial mutations and pervasive epistasis shape the fitness landscape for rifampicin resistance in Pseudomonas aeruginosa. Genetics 186: 1345-1354. [doi: 10.1534/genetics.110.123083]

Payne J.L., Menardo F., Trauner A., Borrell S., Gygli S.M., Loiseau C., et al. (2019) Transition bias influences the evolution of antibiotic resistance in Mycobacterium tuberculosis. PLoS Biol 17(5): e3000265. [doi: 10.1371/journal.pbio.3000265]

Sackman, A.M., McGee, L.W., Morrison, A.J., Pierce, J., Anisman, J., Hamilton, H., Sanderbeck, S., Newman, C., and Rokyta, D.R. (2017) Mutation-Driven Parallel Evolution during Viral Adaptation. Mol. Biol. Evol. 34:3243-3253. [doi: 10.1093/molbev/msx257]

Schluter, D. (1996) Adaptive radiation along genetic lines of least resistance. Evolution, 50:1766-1774. [doi: 10.1111/j.1558-5646.1996.tb03563.x]

Steppan, S.J., Phillips, P.C., and Houle, D. (2002) Comparative quantitative genetics: evolution of the G matrix. Trends in Ecology & Evolution, 17:320-327. [doi: 10.1016/S0169-5347(02)02505-3]

Stoltzfus, A. and McCandlish, D.M. (2017) Mutational Biases Influence Parallel Adaptation, Molecular Biology and Evolution 34:2163–2172, [doi: 10.1093/molbev/msx180]

Storz J.F., Natarajan C., Signore A.V., Witt C.C., McCandlish D.M. and Stoltzfus A. (2019) The role of mutation bias in adaptive molecular evolution: insights from convergent changes in protein function. Phil. Trans. R. Soc. B [doi: 10.1098/rstb.2018.0238]

The Evolutionary Synthesis: Perspectives on the Unification of Biology E. Mayr and W.B. Provine eds Harvard University Press, Cambridge MA, USA (1980)

Yampolsky, L.Y., and Stoltzfus, A. (2001) Bias in the introduction of variation as an orienting factor in evolution. Evolution & development, 3:73-83. [doi: 10.1046/j.1525-142x.2001.003002073.x]

12 comments :

  1. "Infinitesimal variation" ?

    It seems to me that after making the worthy attempt to fit equations to what they thought was going on biologically, some biologists got overly enamored with the mathematics and let that guide their intuitions of what was happening in the real world

    ReplyDelete
  2. I have a feeling I dont understand this, because it doesnt seem like it should be an issue
    For decades we've known that eusocial insects became eusocial because of a detail of how their sex chromosomes operated. This implies that if you went back far enough in time you could in principle find 2 ascocial species of insects that were similar and closely related but with 2 different ways of determining sex. One would inevitably evolve eusociality, the other wouldn't. This implies that the adaptive landscape for the 2 species would be different, and that they'd have a different set of adaptive and detrimental mutations. .... and, I would think, a mostly overlapping with slight differences..set of neutral mutations.

    ReplyDelete
    Replies
    1. For decades we've known that eusocial insects became eusocial because of a detail of how their sex chromosomes operated.
      Well, no we haven't. That doesn't explain why most haplodiploid insects are not eusocial, and it doesn't explain why a few non-haplodiploid insects are eusocial. Haplodiploidy would seem to make eusociality somewhat more likely or accessible, but that's all. Nor does it seem to have anything to do with what set of mutations they would have, though it may have something to do with what mutations would be advantageous. But so does any environmental context, whether external or internal.

      Delete
    2. OK, I thought the distinction with social/asocial vs sex determination was complete. I guess they have alot more karyotypes now. Still..making it more likely still makes the point.

      As for the rest the point I was trying to make is the although haplo-diploploidy has nothing to do with sociality it certainly seems to drive a species in that direction and for the reasons expressed by that famous geneticist (?) "I'd die for my brother or 2 cousins etc etc..."?
      I thought that was what the attack paper was disputing but I must be missing subtleties here. I'll try to read one of the review papers listed and look up a few terms.

      Delete
    3. When Bill Hamilton discovered the haplo-diploidy relationship in kin selection in the early 1960s, he didn't jump up and down and lay claim to a new law of evolutionary biology that overturned all older theory and contributions. Hamilton, different from others, knew he was able to see farther because he had had the privilege of standing on the shoulders of giants.

      Delete
    4. @W.Benson

      Hamilton, like most evolutionary biologists at that time, was working within the paradigm that (almost) everything in evolution could be explained by some form of natural selection. He developed theories that could explain seemingly anomalous results and salvage the idea that they were still due to natural selection. He was defending the dominant paradigm, not trying to overthrow it.

      Delete
    5. W.Benson is channeling the reactionaries for us, offering a perfect example for discussion in Rhetoric 101. There is no scientific argument here, no substantive claim, no matter of logic or evidence. All he has to offer to this discussion is "Sit down, boy, and shut up. Listen to your betters."

      Delete
  3. I think that these posts will be very informative and interesting, even in the case that we do not share the same opinion about modern evolutionary biology. Thanks for sharing.

    ReplyDelete
  4. Yes question your critics but please no REACTIONARY term. that was never a real thing in the old days. its day has passed as a defining of motives of opponents.
    "the establishment of a fundamental new principal" Has it been established? if something short of this then its not established.
    As a creationist i see all these mutation concepts as simply math ones. Just WHAT can mutations do to organize biology. if one accepts mutations can do anything then it is just math that proves the trail. So this is all options in math.
    The "culture war" four points are instructive. The 'abuse of history" sure makes conclusions suspect in evolutionary biology. makes conclusions as very influence by people and not just pure scientific methodology. Creationists bump into this.

    ReplyDelete
  5. Be gentle, I'm a layperson.

    Looking at the Payne 2019 paper and thinking about transition/transversion bias: Would transitions have a greater likelihood of being synonymous than transversions? If so, would this lead to an effect that one adaptively successful transition mutation implies others, whereas one adaptively successful transversion mutation does not?

    ReplyDelete
    Replies
    1. Almost every transition in the 3rd position of a codon is synonymous, so yes, the fraction of possible synonymous transitions is higher. But this is not the basis for the bias seen by Payne, et al. The main block of data concerns non-synonymous changes. Then they also looked at non-coding changes and found transition bias. So, synonymous changes are not relevant to the observed bias.

      Delete
    2. Thanks, leaves me with more questions to try to learn answers to, which is always fun.

      Delete