4 ms·

The Neutral Theory in Light of Natural Selection

51 points by ArtWomb 8y ago

kurthr 8y ago
Abstract: We evaluate the explanatory power of the neutral theory of molecular evolution, 50 years after its introduction by Kimura. We argue that the neutral theory was supported by unreliable theoretical and empirical evidence from the beginning, and that in light of modern, genome-scale data, we can firmly reject its universality. The ubiquity of adaptive variation both within and between species means that a more comprehensive theory of molecular evolution must be sought.
NPMaxwell 8y ago
Wikipedia explanation of Neutral Theory: "most evolutionary changes and most of the variation within and between species is not caused by natural selection but by genetic drift of mutant alleles that are neutral. A neutral mutation is one that does not affect an organism's ability to survive and reproduce. The neutral theory allows for the possibility that most mutations are deleterious, but holds that because these are rapidly removed by natural selection, they do not make significant contributions to variation within and between species at the molecular level. Mutations that are not deleterious are assumed to be mostly neutral rather than beneficial. In addition to assuming the primacy of neutral mutations, the theory also assumes that the fate of neutral mutations is determined by the sampling processes described by specific models of random genetic drift." https://en.wikipedia.org/wiki/Neutral_theory_of_molecular_evolution https://en.wikipedia.org/wiki/Neutral_theory_of_molecular_ev...
- nwatson 8y ago
  The article is long and I don't have time to read this, but the article seems to imply that "neutral" mutations aren't driving the split between species. I would have thought though that for a given species in a given timeframe facing many-dimensional gradient pressure for "fitness" that: - few outlandishly positive mutations would crop up - natural selection would quickly weed out deleterious mutations (the article addresses this) - over time a very large number of "neutral" mutations and therefore neutral variations within the species would accumulate since their effect is completely orthogonal to the current pressures the species faces (the article implies that this is not relevant for evolution, however I think it would be extremely relevant as I describe following ...) The "neutral" mutations would form the "library" of variations that would become relevant at a LATER time when conditions change for a subset of the population of the species and a different gradient becomes relevant. Some of the previously neutral changes would become deleterious in the new setting, while others would provide outright advantages. It is these surviving mutations that would give the sub-group of the original species its new identity. Conclusion: I would think that the mutations that drive the new species' initial offshoot would come from the inherent variation built up by the prior neutral mutations ... i.e., the evolution was already primed by the prior library of previously neutral mutations and not something that happened solely after isolation and changes in conditions. EDIT: wording
ArtWomb 8y ago
It's amazing to think first-principles mechanisms which trigger mutations at the genetic level are still shrouded in mystery. Cold War era research sought to establish effects on radiation exposure in pursuit of the question: can our species survive the fallout of a general nuclear exchange. More recent imaging techniques have actually been able to visualize the nucleotide "wobble" that results in replication errors. But the environmental information that directly codes for beneficial advantage seems to be a deep secret. And one that could benefit from active inquiry ;)
- simonh 8y ago
  >But the environmental information that directly codes for beneficial advantage seems to be a deep secret. How can any natural process directly code for beneficial advantages? Isn’t whether or not it turns out beneficial determined by selective pressures, not the mechanism causing genetic change?
nonbel 8y ago
>"Furthermore, the neutral theory is not merely a neutral model, to be used as a null hypothesis against which more interesting hypotheses can be tested." What possible scientific purpose would this serve? Why not just test the interesting hypothesis and leave out the strawman?
- ssivark 8y ago
  When the model predictions are deterministic and easy to observe, the null hypothesis is not needed. When model predictions are statistical, then, you want to answer the question: how "likely" is this model, given the data? Answering that with a probability (as would be done in a Bayesian approach) is hard because there is no natural "space of models" over which to compute that probability. So, in the frequentist approach, you check whether the data is sufficiently more likely to have resulted from your model, rather than the null hypothesis model. The "sufficiently more likely" is quantified by a p-value.
  - nonbel 8y ago
    >"When model predictions are statistical, then, you want to answer the question: how "likely" is this model, given the data?" Huh? Here is a model with statistical predictions. Its a model of flipping a fair coin 20 times a day for 365 days: p = 0.5 nDays = 365 nPerDay = 20 dat = rbinom(nDays, nPerDay, p) hist(dat) https://image.ibb.co/bwuYCT/flips.png https://image.ibb.co/bwuYCT/flips.png Here is how I compare theoretical to actual results: > mean(dat) [1] 10.06575 > var(dat) [1] 5.479181 > > nPerDay*p [1] 10 > nPerDay*p*(1-p) [1] 5 >"So, in the frequentist approach, you check whether the data is sufficiently more likely to have resulted from your model, rather than the null hypothesis model. The "sufficiently more likely" is quantified by a p-value." A p-value says something about the fit of the model used to calculate it. It doesn't say anything about any other models.
    - ssivark 8y ago
      Your analysis tells nothing about whether the model of fair coin being flipped is confirmed. You see a deviation of 0.06 in the mean and 0.47 in the variance. Is that negligible? Imagine you had much less data i.e. nDays=5 — What if your sample had a mean of 10.6 and a variance of 4.8? Could you confirm or rule out the fair coin model? Or confirm/ rule out the 0.6-heads-0.4-tails model? Within the frequenting framework, you’d need a null hypothesis to answer that question. Regarding p-value, I agree I could have phrased my comment better. In the frequentist framework, there is no way to confirm a model. The only way to arrive at a model is to falsify all “simpler” models one can think of, that might possibly have explained the data.

51 points by ArtWomb 8y ago

Abstract: We evaluate the explanatory power of the neutral theory of molecular evolution, 50 years after its introduction by Kimura. We argue that the neutral theory was supported by unreliable theoretical and empirical evidence from the beginning, and that in light of modern, genome-scale data, we can firmly reject its universality. The ubiquity of adaptive variation both within and between species means that a more comprehensive theory of molecular evolution must be sought.

NPMaxwell 8y ago

Wikipedia explanation of Neutral Theory: "most evolutionary changes and most of the variation within and between species is not caused by natural selection but by genetic drift of mutant alleles that are neutral. A neutral mutation is one that does not affect an organism's ability to survive and reproduce. The neutral theory allows for the possibility that most mutations are deleterious, but holds that because these are rapidly removed by natural selection, they do not make significant contributions to variation within and between species at the molecular level. Mutations that are not deleterious are assumed to be mostly neutral rather than beneficial. In addition to assuming the primacy of neutral mutations, the theory also assumes that the fate of neutral mutations is determined by the sampling processes described by specific models of random genetic drift." https://en.wikipedia.org/wiki/Neutral_theory_of_molecular_evolution https://en.wikipedia.org/wiki/Neutral_theory_of_molecular_ev...

nwatson 8y ago
The article is long and I don't have time to read this, but the article seems to imply that "neutral" mutations aren't driving the split between species. I would have thought though that for a given species in a given timeframe facing many-dimensional gradient pressure for "fitness" that: - few outlandishly positive mutations would crop up - natural selection would quickly weed out deleterious mutations (the article addresses this) - over time a very large number of "neutral" mutations and therefore neutral variations within the species would accumulate since their effect is completely orthogonal to the current pressures the species faces (the article implies that this is not relevant for evolution, however I think it would be extremely relevant as I describe following ...) The "neutral" mutations would form the "library" of variations that would become relevant at a LATER time when conditions change for a subset of the population of the species and a different gradient becomes relevant. Some of the previously neutral changes would become deleterious in the new setting, while others would provide outright advantages. It is these surviving mutations that would give the sub-group of the original species its new identity. Conclusion: I would think that the mutations that drive the new species' initial offshoot would come from the inherent variation built up by the prior neutral mutations ... i.e., the evolution was already primed by the prior library of previously neutral mutations and not something that happened solely after isolation and changes in conditions. EDIT: wording

ArtWomb 8y ago

It's amazing to think first-principles mechanisms which trigger mutations at the genetic level are still shrouded in mystery. Cold War era research sought to establish effects on radiation exposure in pursuit of the question: can our species survive the fallout of a general nuclear exchange. More recent imaging techniques have actually been able to visualize the nucleotide "wobble" that results in replication errors. But the environmental information that directly codes for beneficial advantage seems to be a deep secret. And one that could benefit from active inquiry ;)

simonh 8y ago
>But the environmental information that directly codes for beneficial advantage seems to be a deep secret. How can any natural process directly code for beneficial advantages? Isn’t whether or not it turns out beneficial determined by selective pressures, not the mechanism causing genetic change?

nonbel 8y ago

>"Furthermore, the neutral theory is not merely a neutral model, to be used as a null hypothesis against which more interesting hypotheses can be tested." What possible scientific purpose would this serve? Why not just test the interesting hypothesis and leave out the strawman?

ssivark 8y ago
When the model predictions are deterministic and easy to observe, the null hypothesis is not needed. When model predictions are statistical, then, you want to answer the question: how "likely" is this model, given the data? Answering that with a probability (as would be done in a Bayesian approach) is hard because there is no natural "space of models" over which to compute that probability. So, in the frequentist approach, you check whether the data is sufficiently more likely to have resulted from your model, rather than the null hypothesis model. The "sufficiently more likely" is quantified by a p-value.
- nonbel 8y ago
  >"When model predictions are statistical, then, you want to answer the question: how "likely" is this model, given the data?" Huh? Here is a model with statistical predictions. Its a model of flipping a fair coin 20 times a day for 365 days: p = 0.5 nDays = 365 nPerDay = 20 dat = rbinom(nDays, nPerDay, p) hist(dat) https://image.ibb.co/bwuYCT/flips.png https://image.ibb.co/bwuYCT/flips.png Here is how I compare theoretical to actual results: > mean(dat) [1] 10.06575 > var(dat) [1] 5.479181 > > nPerDay*p [1] 10 > nPerDay*p*(1-p) [1] 5 >"So, in the frequentist approach, you check whether the data is sufficiently more likely to have resulted from your model, rather than the null hypothesis model. The "sufficiently more likely" is quantified by a p-value." A p-value says something about the fit of the model used to calculate it. It doesn't say anything about any other models.
  - ssivark 8y ago
    Your analysis tells nothing about whether the model of fair coin being flipped is confirmed. You see a deviation of 0.06 in the mean and 0.47 in the variance. Is that negligible? Imagine you had much less data i.e. nDays=5 — What if your sample had a mean of 10.6 and a variance of 4.8? Could you confirm or rule out the fair coin model? Or confirm/ rule out the 0.6-heads-0.4-tails model? Within the frequenting framework, you’d need a null hypothesis to answer that question. Regarding p-value, I agree I could have phrased my comment better. In the frequentist framework, there is no way to confirm a model. The only way to arrive at a model is to falsify all “simpler” models one can think of, that might possibly have explained the data.