Talk:Mutation bias

To do

 * Add table 2 from open-access article by Katju & Bergthorsson (https://academic.oup.com/gbe/article/11/1/136/5209700), which shows numerous examples of titv bias and indel bias
 * add more examples for GC-AT bias from https://academic.oup.com/gbe/article/11/1/136/5209700. Arabidopsis also shows a bias toward AT
 * fix issues with male mutation bias section.
 * this is over-written with too many subsections. it can be rewritten in one section so that it does not seem to dominate the other sections
 * add references to latest work from Przeworski and others
 * male-driven evolution is an evolutionary hypothesis. it is not the same thing as mutation bias which is a genetic hypothesis.
 * Haldane 1935 does not propose the bias hypothesis. This is in Haldane 1947.
 * the mechanism is not established. It was long believed that this is a matter of counting replications but that hypothesis is no longer secure — Preceding unsigned comment added by Dabs (talk • contribs) 18:18, 12 June 2023 (UTC)
 * add STR biases section. see Bhargava and Fuentes (doi: 10.1007/s12033-009-9230-4), particularly their comments on the PS/PM model, for references to sources documenting mutational biases.
 * points to explain. Mutation bias is not the same thing as mere heterogeneity.  For instance, consider a genome in which the 3 possible changes to each nucleotide have rates drawn randomly from {u1, u2, and u3}, where these rates are unequal: in this case, there is heterogeneity but no mutation bias.
 * mention that there is considerable heterogeneity in nt mutation rates that has not been assigned to any bias but which presumably reflects some kind of context-dependence. Hodgkinson, et al. http://www.ncbi.nlm.nih.gov/pubmed/21969038
 * points to explain: some biases are directional (they have effects that accrue on some state variable) and others are not. titv bias is not directional. GC-AT bias and indel bias are directional.  This is explained by Stoltzfus 2019.
 * quantitatively varying traits vs. discretely changing traits. Charlesworth 2013 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3730922/. See Uricchio on human variation.
 * add phenotypic biases section. see review by Uller, et al on developmental biases
 * ask to use tree from http://www.sexchrlab.org/male-mutation-bias to show distribution of male mutation bias

Distinguish from patterns of phenotypic variation
In the "related concepts" section, it would be helpful to add some text such as the following. Dabs (talk) 19:50, 29 August 2019 (UTC)

Discussions of patterns of phenotypic variation frequently focus on standing variation, i.e., the variation currently in a population, rather than the generation of new variation by mutation and altered development. By contrast, mutation bias is a reference to the generation of variants.

In quantitative evolutionary genetics, the G matrix represents variances and covariances for quantitative traits in standing variation, whereas the M matrix represents the contribution of new variation due to mutation. The classical versions of G and M cannot have unidimensional biases because they consist of variances and covariances only, with an implied mean of 0. However, it is possible to develop theories for the evolution of quantitative traits subject to mutation bias (Waxman and Peck, others).