More Recent Comments

Thursday, May 11, 2023

Chapter 7: Gene Families and the Birth & Death of Genes

This chapter describes gene families in the human genome. I explain how new genes are born by gene duplication and how they die by deletion or by becoming pseudogenes. Our genome is littered with pseudogenes: how do they evolve and are they all junk? What are the consequences of whole genome duplications and what does it teach us about junk DNA? How many real ORFan genes are there and why do some people think there are more? Finally, you will learn why dachshunds have short legs and what "The Bridge on the River Kwai" has to do with the accuracy of the human genome sequence.

Click on this link to see more.

Gene Families and the Birth and Death of Genes


Unknown said...

Just to say, I appreciate you broadcasting the chapter summaries.

João said...

Reading chapter 7 of your amzing book. You say the following:

"Mutationism is just one of many ideas that emphasize the randomness and
unpredictability of evolution, and it fits nicely into the view that genomes are sloppy and not exquisitely designed like a Swiss watch."

Honest question: isn't the works of Arlin and other researches on mutation bias making adaptive evolution more predictable?

Maybe is all about scales and how one defines mutationism?

Great book, Larry! I'm really happy it was finally published.

Larry Moran said...


Mutation bias can determine the direction of evolution but that's not the same as "adaptive evolution." For example, a given mutation bias may allow you to predict that the GC content of a genome will increase or decrease but most of those changes will be due to fixation by random genetic drift.

The original supporters of the Modern Synthesis viewed mutation as irrelevant. They thought that that evolution was mostly due to natural selection operating on a large source of existing alleles such that whenever a adaptation was necessary the relevant allele would be available in the population. You can see this idea in most of Dawkins' writing.

The mutationists argue against this predictability and this form of adaptationism. They argue that the direction of evolution depends to a large extent on the availability of particular mutations that just happen to occur. In Nei's version, for example, accidental gene duplications play a large role.

Take the expansion of Alu sequences in primate genomes, for example. That's just the result of some mutation that converted a 7SL RNA to a transposon but it had significant consequences on the genome composition of primates. It's difficult to explain that if you adhere to the traditional Modern Synthesis view of evolution.

João said...


Yes, I know that mutation bias cannot be conflated with adaptive evolution. And I'm no defender of "why the orthodoxy is true". However, mutation bias can influence adaptive evolution.

"How do mutational biases influence the process of adaptation? A common assumption is that selection alone determines the course of adaptation from abundant preexisting variation. Yet, theoretical work shows broad conditions under which the mutation rate to a given type of variant strongly influences its probability of contributing to adaptation."

From Cano et al. 2022

Arlin is co-author to this paper, and here is his brief commentary about it:

And here we have Arlins peaking of mutation-biased adaptation:

Do I misundarstand what Arlin says, or is he claiming that the process of adaptation can indeed be influenced by mutation bias?

Off Topic:

Given Nei's recent passing, will you write an eulogy, emphasizing his main contributions?

Thanks, Larry. Finished chapter 7. And I keep reading!

Joe Felsenstein said...

They argue that the direction of evolution depends to a large extent on the availability of particular mutations that just happen to occur.
That is too sweeping, and I hope they don't really say that. The exact implementation of changes may well depend on what particular mutations occur. But when a population is in a situation where increased adaptation to (say) cold would increase fitness, I doubt that the occurrence of particular mutations will bring about decreased adaptation to cold.

Larry Moran said...


There are many different ways to adapt to cold and the exact pathway will depend on the available alleles in the population. That's part of what I mean by the "direction" of evolution. For example, some fish have evolved antifreeze proteins whereas other have followed a different pathway. Some have not adapted at all so they avoid the Artic and Antarctic oceans.

But I'm also referring to new adaptations such as the evolution of trichromatic vision in primates. That depended on the chance duplication of one of the opsin genes and its divergence from the parent gene so it could detect different wavelengths of light. This mutation didn't occur in most mammals so they only have dichromatic vision. Primates have evolved in a different direction.

You are looking at a particular adaptation in response to some environmental pressure but there's more to the direction of evolution than that. Nei's latest book (Mutation-Driven Evolution) has chapters on the role of gene duplications and polyploidization and a very interesting chapter on the role of mutation in speciation.

Arlin said...

@Larry, who are the "mutationists" here? I haven't got the book yet. The early geneticists did not embrace this label, nor did Nei, nor do I. Currently, this word is radioactive, due to its strawman meaning in Synthesis propaganda. But it might be possible to recover a legitimate meaning:

As @João points out, "theoretical work shows broad conditions under which the mutation rate to a given type of variant strongly influences its probability of contributing to adaptation." This effect has been demonstrated empirically using data from experimental and natural adaptation.

@joe, the scenario that you describe is perhaps the sole case in which one expects no meaningful influence of mutational preferences: a single quantitative trait with abundant infinitesimal variation both up and down, where only one of these directions is adaptive. Everything we know suggests that, in this specific scenario, evolution will go up or down depending on which direction is beneficial, with only a tiny displacement from the optimum due to mutation bias (see Waxman and Welch, or Charlesworth or Zhang and Liu).

More generally, however, evolution is an exploration in a multi-dimensional space, so that there are many ways to go up. The theory of arrival biases provides a way to understand how and why evolution tends to go up in ways that are mutationally favored.

The wikipedia page has links to sources and a good summary of theory, evidence, and implications:

Larry Moran said...


I'm using "mutationist" as a way of identifying those who advocate mutation-driven evolution. I call this "mutationism" in my book and I point out that you and Nei are proponents of this view.

I realize that that term "mutationism" has other meaning dating back to the early twentieth century but, as Nei points out, that view was not true mutationism in the sense that he uses the term today. He refers to his view as a "new form of mutationism" but I don't think he used the term 'mutationist."

Would you prefer some other label to identify those who share your point of view?

Arlin said...

@Larry thanks for the clarification. I need to re-read Nei. I looked back at my notes and realized that he has an explicit discussion of mutationism and neo-Darwinism before referring to "neo-Mutationism" as a label for "mutation-driven" evo.

Before talking about what labels are appropriate, we need to be clear about what, precisely, are the positions to be labeled. Are these research programs or theories? What are the specifics? That what the blog is about:

A decade ago, I read and re-read Nei's book, and wrote countless drafts of a book review, trying to capture his views, which IMHO are a mess philosophically. Throughout the book he argues that "since phenotypic evo is ultimately controlled by DNA and RNA molecules, both molecular and phenotypic evo must be primarily caused by mutation." (p. 9) Note that this claim purports to follow from upward causation (genes to phenotypes) alone, without knowing anything whatsoever about how the process of evolution emerges from population dynamics. If this is what "mutation-driven" means, then it's a bad arg and I would reject it utterly.

But one could also argue from Nei's book that "mutation-driven" is an explanatory position, based on the way that "driving" is used throughout quantitative sciences for a factor that is strongly predictive. If an economist says that global droughts are driving food prices, this does not mean that droughts are a mechanical force pushing on prices, but that this factor is predictive.