Sex with no regrets: How sexual reproduction uses a no regret learning algorithm for evolutionary advantage

Abstract

The question of 'why sex' has long been a puzzle. The randomness of recombination, which potentially produces low fitness progeny, contradicts notions of fitness landscape hill climbing. We use the concept of evolution as an algorithm for learning unpredictable environments to provide a possible answer. While sex and asex both implement similar machine learning no-regret algorithms in the context of random samples that are small relative to a vast genotype space, the algorithm of sex constitutes a more efficient goal-directed walk through this space. Simulations indicate this gives sex an evolutionary advantage, even in stable, unchanging environments. Asexual populations rapidly reach a fitness plateau, but the learning aspect of the no-regret algorithm most often eventually boosts the fitness of sexual populations past the maximal viability of corresponding asexual populations. In this light, the randomness of sexual recombination is not a hindrance but a crucial component of the 'sampling for learning' algorithm of sexual reproduction.

Gene copy number variations as signatures of adaptive evolution in the parthenogenetic, plant‐parasitic nematode Meloidogyne incognita

“…For over a century, the paradigm has been that the lack of recombination associated with asexual reproduction should reduce additive genetic variance and produce clonal progenies, thus resulting in poor capacities to generate new combination of genotypes (Crow, ; Edhan, Hellman, & Sherill‐Rofe, ; Song, Scheu, & Drossel, ). In this context, adaptation to changing environmental conditions represents a challenge to parthenogenetic organisms.…”

Sex solves Haldane’s dilemma

“…Muller (1932Muller ( , 1964) developed Fisher's idea and concluded that recombination among many loci could allow a sexual population to evolve "hundreds to millions of times faster" than a comparable asexual population. This same idea, of an increasing advantage for recombination as the number of loci under selection increases, has been echoed in later papers (e.g., Crow and Kimura 1965;Otto and Barton 2001;Colegrave 2002;Iles et al 2003;Park and Krug 2013;Edhan et al 2017). Park and Krug (2013) have shown that the advantage of recombination is proportional to L, where L is the number of loci under selection.…”

“…In a genetically polymorphic sexual population multilocus genotypes are highly ephemeral. This is because existing genotypes are merely a small random sample of a much larger set of possible genotypes (Ewens 1972;Edhan et al 2017). In this study, we assumed-for computational convenience-that there were only 100 biallelic loci.…”

Sex Solves Haldane’s Dilemma

“…Muller (1932, 1964) developed Fisher’s idea and concluded that recombination among many loci could allow a sexual population to evolve “hundreds to millions of times faster” than a comparable asexual population. This same idea of an increasing advantage for recombination as the number of loci under selection increases, has been echoed in later papers (e.g., Crow and Kimura 1965, Otto and Barton 2001, Iles et al 2003, Park and Krug 2013, Edhan et al 2017). Park and Krug (2013) have shown that the advantage of recombination is proportional to L, where L is the number of loci under selection.…”

“…But when we consider many loci simultaneously, then each genotype is effectively unique. This is because existing genotypes are merely a small random sample of a much larger set of possible genotypes (Ewens 1972, Edhan 2017). Human populations provide a familiar example: every human individual is genetically distinct, not only from the billions of other individuals within the current population, but also from individuals in past and future populations.…”