4.1 Sources, and where an allele waits #
4.1.1 Mutation, recombination, and gene flow #
Selection and drift have nothing to sort until individuals differ heritably. Mutation creates new sequence. Recombination, by independent assortment and crossing over, creates new combinations. Gene flow imports alleles that already exist elsewhere. How a polymerase errs, how a crossover is resolved, and how a plasmid moves are Molecular Biology, Genetics, and Microbiology. The population fact is enough: mutation is the only source of genuinely new alleles; recombination multiplies genotypes; gene flow can add an allele faster than local mutation if a neighbour already has it.
Most mutations that change a protein are deleterious. Many more, especially silent and non-functional non-coding changes, are neutral or nearly so. A minority are beneficial in the current environment. “Random” means random with respect to utility, not equally likely at every base. Hotspots are chemistry, not aim.
The usual fate of a new mutant is loss. In a diploid population of size N a unique neutral allele, at frequency 1/(2N), has probability 1/(2N) of eventual fixation. Deleterious alleles are lost faster. Beneficial alleles are usually lost too, while they are still rare (Chapter 6). Visible polymorphism is the residue: alleles protected by balancing selection, alleles in transit, and neutral alleles drifting in the middle.
4.1.2 Standing variation and new mutation #
Standing variation is polymorphism already segregating. New mutation appears after a challenge begins. Adaptation uses both. A new beneficial mutation must survive the first few generations, when one unlucky draw eliminates it, and it sits on a single haplotype. An allele already at moderate frequency has survived that stage and usually sits on several haplotypes. If the environment then favours it, the response can be fast, and the sequence signature is blurrier than a classical sweep from one new copy (Chapter 6).
Large populations produce more new mutants per generation. Small populations must mostly use what they hold, and drift has left them holding less. Polygenic traits often respond by small frequency shifts at many standing loci rather than by fixation of one new mutation. The breeder’s equation still gives the response. It does not say whether the heritability was old or new.
4.2 The Hardy–Weinberg null #
4.2.1 The statement used here #
Genetics derives the Hardy–Weinberg proportions. These notes use the result. At an autosomal locus with two alleles at frequencies p and q, p + q = 1, if there is no selection, no mutation, no migration, and no drift, and if mating is random, genotype frequencies become p², 2pq, and q² in one generation, and allele frequencies do not change. That is Hardy–Weinberg equilibrium.
Break one assumption and the conclusion that depends on it fails. Non-random mating rearranges genotypes and need not change allele frequencies (Chapter 5). Selection, mutation, migration, and drift can change allele frequencies. For one autosomal locus, a single round of random mating restores the genotype proportions even if the previous generation was not in them, provided the other forces are absent. Two loci together, and sex linkage, do not behave so simply. Those cases stay in Genetics.
A population at p², 2pq, q² has not been shown to be free of selection. Some selection leaves frequencies close to those proportions, and one generation is a weak test. Agreement at one locus says nothing about other loci. The equilibrium is the baseline. Change in allele frequency is a departure from what the baseline would have kept constant. A departure in genotype frequencies, with p unchanged, is a mating or structure problem until shown otherwise.
4.2.2 What moves when frequencies move #
Adults produce gametes. With Mendelian segregation and no selection, gamete frequencies equal adult allele frequencies. Random fusion rebuilds zygotes at p², 2pq, q². If selection then acts on survival, each genotype is weighted by its fitness and divided by mean fitness, and the adults who breed are no longer in those proportions. The allele frequency in their gametes has moved. Chapters 5 and 6 take that movement apart by cause.
Dominance, frequency, and fitness are different words. Dominance is a relation between genotype and phenotype. Frequency is a count. Fitness is reproductive success. A recessive allele can be common, harmful, and still increase by drift in a particular small population.
4.3 Neutral and nearly neutral change #
4.3.1 Kimura’s claim #
Motoo Kimura’s neutral theory (1968; developed by Kimura, and argued independently from proteins by King and Jukes) says that most substitutions fixed between species at the molecular level are the result of drift acting on neutral or nearly neutral alleles. It does not say that most mutations are neutral. Most mutations that alter function are deleterious and are removed by purifying selection. They never become substitutions. Among mutations that are fixed, neutral ones predominate, because deleterious ones are eliminated and beneficial ones are rare.
For a strictly neutral allele the substitution rate equals the neutral mutation rate. Population size cancels: a larger population has more mutants and a smaller fixation probability per mutant. That is why a molecular clock can exist. Clocks are not metronomes. Mutation rate per year depends on generation time and germline biology, and it differs among lineages and among kinds of site. A clock date is an estimate, calibrated often against fossils, with error. It is not itself a fossil date. Chapter 9 uses geological dates where it has them.
Nearly neutral theory, especially Ohta’s, covers alleles with selection coefficients small relative to 1/Ne. When |s| is much smaller than 1/Ne, drift dominates. A small population therefore fixes some slightly deleterious alleles that a large population would remove. Phenotypic evolution in the same species can still be dominated by selection. Neutral theory is a claim about molecular substitutions, not a claim that beaks and drug resistance evolve by chance.
4.3.2 dN/dS without a methods manual #
In a coding sequence, a synonymous change alters the codon and not the amino acid. A nonsynonymous change alters the amino acid. The ratio of their rates, dN/dS or ω, compares them. ω below 1 means amino-acid changes are removed more often than silent changes: purifying selection, and the usual result for a functional protein. ω about 1 is consistent with neutrality of amino-acid change. ω above 1 means amino-acid changes have been fixed faster than silent ones: positive selection. A whole-gene average above 1 is uncommon, and it is strong evidence when it is real.
Allopatric Speciation With Reinforcement Schematic
A gene-wide ω below 1 does not mean that no site was positively selected. Purifying selection elsewhere pulls the average down. Positive selection is often limited to a few sites or one branch. Synonymous sites are not perfectly neutral either: codon use, splice signals, and RNA structure can constrain them. Reading ω is an idea about rates. Estimating it — alignment, models, the choice of sites and branches — is a method, and it is not taught here. Alignment, on which any such comparison depends, belongs to Bioinformatics. Most proteins show purifying selection. A minority of sites show positive selection. Both are compatible with the neutral theory, because one claim is about what gets fixed and the other is about what gets removed.
4.4 Key terms, summary, and self-check #
Key terms. Mutation; recombination; gene flow; standing variation; Hardy–Weinberg equilibrium; neutral theory; molecular clock; nearly neutral theory; purifying selection; dN/dS.
Summary. New alleles come from mutation, new genotypes from recombination, new arrivals from gene flow, and almost every new neutral allele is eventually lost. Hardy–Weinberg frequencies p², 2pq, and q², with constant allele frequencies, are the null when selection, mutation, migration, and drift are absent and mating is random. Neutral theory says most molecular substitutions are fixed by drift, at a rate equal to the neutral mutation rate. ω below 1 is purifying selection and is the usual gene-wide result; ω above 1 is positive selection and is rarely true of an entire gene even when a few sites are under it.
Self-check.
1. Which of mutation, recombination, and gene flow can supply an allele the species has never had?
2. State the Hardy–Weinberg genotype frequencies and the assumptions that hold allele frequencies constant. Why does one round of random mating restore the genotype frequencies at one autosomal locus?
3. Most mutations that change a protein are deleterious, yet most substitutions between species are neutral. How can both be true?
4. Why is the neutral substitution rate independent of population size? Why does a nearly neutral allele care about Ne?
5. A gene has ω well below 1. Why can a few of its sites still have been fixed by positive selection?
—
—
—