Re: latent mutation

From: Chris Taylor (Christopher.Taylor@man.ac.uk)
Date: Wed 07 May 2003 - 11:52:26 GMT

  • Next message: Keith Henson: "Communism as a "religious" meme (renamed thread)"

    Ok here we go :)

    When doing the multiplicative thing with those numbers for reversing a point mutation we should bear a few things in mind. First, when a phenotypic trait 'reverses' it is often due to a compensation by a different gene, rather than the 'fixing' of the mutated one.

    Second, homologous recombination and sex allow the 'fixing' of genes:

    A----o-----\/--------------B C----------/\-------o------D

    If the little 'o's are the point mutations, and the big X thing is the recombination site (a chiasma), then we can perhaps see that two
    'damaged' genes from different parents could recombine in the child to produce one double mutant allele (A->'X'->D) and one 'fixed' one, which was previously lost to the population (C->'X'->B). Then there is a 50:50 that the 'good' copy is the one that gets passed on (for each gamete, ignoring meiotic drive). This is actually one of the major components of the argument for the existence sex (along with parasitism). If you're really eager, look up Muller's ratchet - this is the situation where the fittest class is irredeemably lost because there is no recombination plus sex, and you really are relying on true reverse mutation (which basically never happens) to fix things.

    Thirdly, bearing in mind that all codons (for a single amino acid) are made up of three consecutive nucleotides, an important mutation is the frame shift, where a base insertion/deletion shifts the portion after it
    (in the sense of the direction taken by the translation machinery [5' to 3' in the techie speak]) so that position 1 becomes position 2, 2 becomes 3 (for insetion) or 2 becomes 1, 1 becomes 3 in the previous one etc. for a deletion. These frame shifts are often at the bottom of the most damaging effects of mutation (point mutations are frequently harmless in and of themselves, for various reasons). So if all you have to do is correct the deletion frame shift with an insertion near enough not to cause too much collateral damage (or vice versa), you can
    'reverse' the mutation. This is much more probable than hitting the exact spot.

    As an aside, consider tRNA stem regions (don't worry about the detail) - mutations appear to occur in pairs here (actually one at a time over a shortish time frame). The point being that this is a requirement (given the change in one of the pair) for a precise point mutation, with no mechanism to aid it, to complement the initial change. The 'hard' case I believe (remember these things are translated from genomic DNA just like everything else, and what is paired in the folded tRNA will be many bases apart in the genome so no help there). Yet these paired things occur quite a lot (enough to resolve evolutionary trees quite well, for example). It certainly took a while for me to accept that something so precise and seemingly unlikely could occur at a measurable rate.

    Another aside (then I'll shut up). Against the multiplication of improbablilies, divide through by the number of gametes that fail to develop properly - these too undergo a sort of selection, and the huge numbers of offspring that are produced but rapidly die in many species
    (many insect species, especially flies, have as much as 99 percent progeny-free mortality).

    Oh yeah - Scott's post...

    <snip...>

    > In polyphenism (sometimes called polymorphism) there's not a genetic
    > difference between individuals responsible for difference in phenotypic
    > appearance, but differences in development.

    I saw a nice composite photo of a pine (or maybe it was a spruce) a while ago; when it grew at low altitude you got a big triangular tree, but when it grew at high altitude you got a little shrub, with a little ball shaped lump of foliage. That is phenotypic plasticity (polyphenism
    / morphism) - essentially a genome with 'options'. As is our capacity to learn (in a more general sense where memes affect our physical and mental being).

    > Maybe, if the basis is genetic difference in individuals within a
    > popluation across time, there might be two alleles involved where at one
    > time one form becomes rare, but when the adaptive lanscape shifts back,
    > this rare form becomes a selective advantage and the frequencies
    > eventually shift back to a state where this form becomes common and the
    > other rare. If there's migration from an area where the situation is
    > reversed, an occasional migrant or so could bring their alleles into the
    > population where they have been rare but presently adaptive. This "gene
    > flow" plus mutation could be factors in subsequent population changes
    > (accounting for population size and all that).

    Yeah it's almost impossible to argue that the 'reversed' version of the gene wasn't just lurking at really low frequencies from generation to generation until it found favour (the modelling shows that it is
    *really* hard to get rid of the last dregs). Biston betularia is the textbook example of this (and most of the work was done in Manchester hehe), another good example is the Cain and Sheppard stuff with Cepaea snails under visual predation, in grassland versus woodland.

    Phew. Hang on I definitely remember having a job of some kind...

    Cheers, Chris.

    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
      Chris Taylor (chris@bioinf.man.ac.uk)
      http://bioinf.man.ac.uk/ »people»chris
    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

    =============================================================== This was distributed via the memetics list associated with the Journal of Memetics - Evolutionary Models of Information Transmission For information about the journal and the list (e.g. unsubscribing) see: http://www.cpm.mmu.ac.uk/jom-emit



    This archive was generated by hypermail 2.1.5 : Wed 07 May 2003 - 12:01:11 GMT