Genetic values - mutations with direct effects on fitness

Note

The objects described here are passed to the gvalue parameter when initializing instances of fwdpy11.ModelParams.

In a typical population-genetic model, mutations have direct effects on fitness. Often, this effect is referred to as s, or the “selection coefficient”.

Once we’ve decided on our distributions of effect sizes, we need a way to obtain a diploid’s fitness. For these “standard” population genetic models, we will use fwdpy11.Multiplicative. Instances of this class tell the simulation to calculate the genetic value of an individual using a multiplicative model where the value contributed by each position with a mutation is:

Genotype	AA	Aa	aa
Fitness	\(1\)	\(1+hs\)	\(1 + scaling\times s\)

In this table:

• A refers to the ancestral/non-mutant allelic state

• a is the mutant allelic state

• h is the heterozygous effect of the mutant, the so-called dominance coefficient.

• s is the selection coefficient.

• scaling lets you decide between Fisher, Wright, Haldane, Kimura, etc., when determining the fitness of the mutant homozygote.

The most common values for scaling are 1.0 or 2.0:

import fwdpy11

gvalue = fwdpy11.Multiplicative(scaling=1.0)
gvalue.scaling
gvalue = fwdpy11.Multiplicative(scaling=2.0)
gvalue.scaling

2.0

Note

The scaling parameter interacts with the h parameter for a distribution of effect sizes! (See Distributions of effect sizes.) For example, if scaling = 1.0, then h = 1.0 results in dominant mutations. However, if scaling = 2.0, then h = 1.0 gives co-dominant mutations. In both cases, h = 0.0 generates fully-recessive mutations.