Talk:Selection coefficient

Wiki Education Foundation-supported course assignment

This article is or was the subject of a Wiki Education Foundation-supported course assignment. Further details are available on the course page. Student editor(s): Kwfahey.

Above undated message substituted from Template:Dashboard.wikiedu.org assignment by PrimeBOT (talk) 08:54, 17 January 2022 (UTC)

Untitled

This should probably be merged with evolutionary pressure. Richard001 (talk) 23:51, 13 August 2008 (UTC)

This article makes the claim that "Selection can only act directly on phenotypic differences." It should be noted that this is a contentious issue, with Neo-Darwinians like G. C. Williams and Richard Dawkins arguing that selection is always and only of genes. I don't have a dog in the fight, but i think the article should be edited to acknowledge the fact that there is a reasonable debate currently taking place over the question. —Preceding unsigned comment added by 68.93.74.163 (talk) 04:35, 7 January 2011 (UTC)

Er, no, you've misunderstood; Dawkins would not dispute that selection can only act directly on phenotypic differences. JackAidley (talk) 16:23, 7 January 2011 (UTC)

Introductory paragraph

SELECTION COEFFICIENT: "For example, the lactose-tolerant allele spread from very low frequencies to high frequencies in less than 9000 years since farming with an estimated selection coefficient of 0.09-0.19 for a Scandinavian population. Though this selection coefficient might seem like a very small number (9 to 19 humans were favored in the Scandinavian population)." This makes no sense. Given the definition, this should say that the selection coefficient of the lactose-INtolerant gene is .09-.19, and that this means that people without the gene left 9% to 19% fewer descendants (per generation?).

Also, this and other definitions of selection coefficient state it can range from 0 to 1, yet its definition clearly implies that some genes, including the lactose-tolerance gene being discussed, have negative selection coefficients. Why this discrepancy? Can the lactose-tolerance gene not be said to have a selection coefficient? Philgoetz (talk) 23:07, 15 January 2015 (UTC)

Merge with Fitness (biology)

Fitness (biology) is much further along than this page, and all a selection coefficient is the difference between fitnesses - implying that it can't be understood without a quantitative understanding of fitness anyway. This page could redirect to fitness or to a subsection of fitness. Specifically, I would put a new section "selection coefficient" in between the current sections "relative fitness" and "Change in genotype frequencies due to selection" on the current fitness page. I'm not very savvy with moving/redirecting to a subsection though, and would be happy if someone else volunteered to implement the move. Joannamasel (talk) 20:13, 3 September 2018 (UTC)

I agree. On this page, there's no quantitative explanation. Waylah (talk) 23:17, 30 October 2018 (UTC)

Clarity relative fitness

The notation here for relative fitness is not locally defined (and the fitness (biology) article on Wikipedia comes nowhere close to dealing with this in the lead).

Suppose that there are two genotypes A and B in a population with relative fitnesses $w_{A}$ and $w_{B}$ respectively.

What are these fitness coefficients? Since I don't wish to wade through the thick brambles of fitness (biology), let's instead see what Google says:

• Calculation of Fitness and Selection Coefficient

Relative Fitness (w) is the survival and/or reproductive rate of a genotype (or phenotype) relative to the maximum survival and/or reproductive rate of other genotypes in the population.

Calculate the Relative Fitness (w) of each genotype by dividing each genotype's survival and/or reproductive rate by the highest survival and/or reproductive rate among the 3 genotypes.

By this definition, fitness can never exceed 1.

But not in this neck of the woods, apparently:

Then, choosing genotype A as our point of reference, we have $w_{A}=1$, and $w_{B}=1+s$, where s measures the fitness advantage (s>0) or disadvantage (s<0) of B.

Simply put, the notational convention in use here is not adequately defined. — MaxEnt 18:52, 27 July 2021 (UTC)

I think it depends on the context. In standard cases, the genotype with highest fitness is assigned 1 (w = 1), and all other genotypes have relative fitness values between 0 and 1.

But in other cases when we compare genotypes relative to a chosen reference genotype (rather than the absolute maximum fitness), it can exceed 1 (this indicates "more fit" than the other).

So let's say we compare two genotypes A and B (rather than comparing a genotype against absolute maximum fitness in the population). The reference genotype A is assigned w_A = 1. If genotype B has a fitness advantage over A, it'd have s (selection coeff) > 0. So let's say s of B is 0.20 - this means w_B = 1 + s = 1 + 0.2 = 1.2 (this means that genotype B is 20% more fit than A). But suppose genotype B is less fit than A, it'd have s value < 0, so let's say s = -0.2 then w_B = 1 + (-0.2) = 0.8 (this means genotype B has 80% of the fitness of genotype A - meaning 20% disadvantage in fitness compared to A). --WikiLinuz (talk) 05:22, 3 September 2024 (UTC)