Gene repair and sex

The gene repair theory

The status of the gene repair theory of the evolution of sex is one of the active areas of modern evolutionary biology.

The theory's prominent supporters include Richard Michod and Alexei Kondrashov.

Kondrashov describes the theory in the paper: "Deleterious mutations and the evolution of sexual reproduction" 1988.

Kondrashov's theory is widely believed to require "synergetic mutations" to work.

Problems

The main problem with the gene repair theory is that it faces competition with the 'Red Queen' theory - which explains a similar range of phenomena, but does it rather better.

However, the concern of this essay is narrower than that - it looks at the significance of dysergetic epistasis to the gene repair theory.

Model

In an attempt to throw some light on the area, I wrote a computer simulation of single-celled organisms, reproducing both sexually and asexually.

Directly modelling the effect of mutations on fitness has some realism problems - since it fails to take into account issues such as the fact that late mutations have less impact on reproductive success than early ones - so to avoid this problem I chose to allow mutations to affect mortality.

Increases in mortality automatically compromise reproductive success - by decreasing the organism's remaining lifespan.

I modelled zero epistasis, synergetic epistasis and dysergetic epistasis.

I measured the number of deaths in the population, the average lifespan, the number of mutations each new-born organism carried - and the number of mutations present in each living individual at the end of the run. Mutations were measured since they are a useful proxy for the population's general state of health.

Conclusions so far

My initial conclusions are as follows:

In this model, sexual recombination usually reduces the death rate, reduces the number of mutations in newborns and typical adults, and extends the lifespan of members of the population (as compared to asexual populations).

It does this in a range of cases - including the case of zero epistasis. This doen't directly contradict Kondrashov's model - since in his model mutations are considered to directly affect fitness - whereas in my model (which I hope is more realistic) the mutations affect mortality - and thus can have different effects on fitness depending on the age at which they occur.

Also, it is thought that the balance between synergetic and dysergetic epistasis is matters - since dysergetic epistasis favours asexual reproduction. [e.g. see Mark Ridley's "Evolution" textbook, [Figure 12.6, p. 323 (3rd edition)].

My model suggests that not all types of dysergetic epistasis favour asexual reproduction; and that - that of those that do - the effect appears to be incredibly small.

Therefore, the idea that there is some sort of balance between synergetic and dysergetic epistasis may prove to be worth reexamining. If synergetic interactions favour sex - while finding dysergetic interactions that offer more than a tiny advantage to asexual reproduction seems rather challenging - then the issue of how much dysergetic epistasis occurs may not matter - and the viability of the mutation repair theory would depend only on whether the interactions that favour sex are common enough to offset its costs.

Future work

The most obvious next thing to do with this model is to put the sexual individuals and asexual individuals into the same population - and allow them to compete with one another - to examine the circumstance under which sex is favoured.

Source code

The (public domain) source code is available here.

Raw results

Here's the raw data from one run of the program:

Zero epistasis: effects of mutations on mortality are independent
Asexual experiment
Deaths: 1040406 - Average lifespan:1.9221448
Average mutations at birth:7.3622866
Average mutations per individual at end of run:7.52
Sexual experiment
Deaths: 889552 - Average lifespan:2.2480564
Average mutations at birth:5.746793
Average mutations per individual at end of run:5.03


Dysergetic epistasis: mutations have diminishing effect on mortality
Asexual experiment
Deaths: 1437500 - Average lifespan:1.3912237
Average mutations at birth:11.261751
Average mutations per individual at end of run:16.9
Sexual experiment
Deaths: 1052413 - Average lifespan:1.9002084
Average mutations at birth:1.1641209
Average mutations per individual at end of run:0.67


Dysergetic epistasis: mutations have diminishing effect on mortality
Asexual experiment
Deaths: 1005376 - Average lifespan:1.9890469
Average mutations at birth:0.54179233
Average mutations per individual at end of run:0.28
Sexual experiment
Deaths: 1033338 - Average lifespan:1.9352545
Average mutations at birth:0.55689037
Average mutations per individual at end of run:0.14


Synergetic epistasis: mutations have an increasingly negative effect on mortality
Asexual experiment
Deaths: 60790 - Average lifespan:32.847622
Average mutations at birth:12.590755
Average mutations per individual at end of run:12.82
Sexual experiment
Deaths: 60048 - Average lifespan:33.24732
Average mutations at birth:12.433553
Average mutations per individual at end of run:12.34


Synergetic epistasis: mutations have an increasingly negative effect on mortality
Asexual experiment
Deaths: 602897 - Average lifespan:3.3168137
Average mutations at birth:9.951496
Average mutations per individual at end of run:11.15
Sexual experiment
Deaths: 571718 - Average lifespan:3.4977016
Average mutations at birth:9.355055
Average mutations per individual at end of run:10.14


Tim Tyler | Contact | http://alife.co.uk/