Hi CF,
You say:
Thanks for continuing the gambling analogy; I really don't have the background to debate molecular genetics at a really technical level. I'll leave that to the experts.
Anyway, to continue our analogy - how can some mutations be random and some not?
PB: NRM are introduced on the same spot over and over independent from each other due to mechanism operating at the level of the genome. Such mechanisms have recently been discovered and are described in Caporale's book "Darwin in the genome". One such mechanism is due to an imperfect hairpin model, and results in mutations that are far from randomly introduced: same spot same nucleotide. Besides, for many genes protein driven mechanism have been desribed that introduce mutations in a very specific way. Mutations that are introduced deliberately and dependent on preexistent proteins can not be classified as random, since they are introduced upon a trigger. Whether or not mutations can be introduced as a result of to environmental input remains to be established but we now know that it is absolutely not excluded. See here:
http://
EvC Forum: Doe the climate direct mutations towards the ATP6 gene? -->
EvC Forum: Doe the climate direct mutations towards the ATP6 gene?
Furthermore, mutations introduced through oxidative stress, radiation, etcetera are also not really random since they often are introduced at the same spot. See here:
http://
EvC Forum: More non-random evolution -->
EvC Forum: More non-random evolution
CF: I mean, how could one tell the difference?
PB: I realise that it is difficult to discriminate between random and non random mutations but I have tried to explain it in several threads. For instance here:
http://
EvC Forum: Dr Page's best example of common descent easily --and better-- explained by the GUToB -->
EvC Forum: Dr Page's best example of common descent easily --and better-- explained by the GUToB
PB: It should be realised that comparison of several sequences of (sub)species is the appropriate tool. I discovered NRM in the 1G5 gene since the authors presented the genes of 13 subtypes of Drosophila. But they are also present in the ZFY region and in human mtDNA.
CF: At the roulette wheel, that seems like it's saying "every time the wheel comes up A, it's a non-random mutation - the other three nucleotide results are random mutations."
PB: The most important thing is that the same postion is involved over and over, and usually the same nucleotide. That the rule is not 100% with respect to nucleotide makes it more easy to detect NRM in sequence comparisons.
CF: It seems like you're taking the mutations that are useful or beneficial and saying "these are non-random" even though, at the same location, other, less beneficial mutations occur sometimes as well. It seems like either all mutations must be random, or none are.
PB: It should be noted that at present it is not known whether or not an association exists between the NRM and benefit of the mutations. Although, the offspring of cone snails benefit from the NRM introduced in their toxin genes.
CF: How could one tell the difference between NRM and RM?
PB: That's what Dr Page is going to tell us, since he is using computerprograms that discriminate between RM and NRM.
CF: If RM is happening all over the genome, how do you know that it's not happening at those specific sites?
PB: For RM mutations in mtDNA they are indeed introduced on the same sites. Which makes them in fact NRM.
CF: I mean, to confirm that the mutations are truly non-random you'd have to observe every mutation at that site and discover that every time, the mutation was beneficial. Only then could you determine that the mutation wasn't simply random.
PB: There is to my knowledge no direct association between NRM and benefit of that mutation. All NRM holds is that the position and nucleotide that is being introduced are non-random. But that it will have tremendous impact for phylogenetic analysis is fo sure.
CF: Also, if the mutation is non-random, shouldn't it be possible to predict what the mutation will be before it occurs? Predictability would seem to be the opposite of random.
PB: I think that for certain NRM the possibility exists that we can actually predict where they will be introduced as soon as we elucidate the underlying mechanism. It should be noted that high levels of radiation introduced mutations on the same spot in the mtDNA over and over. So eventually we will be able to exactly determine where the next mutaion will be introduced (within a certain probability range). But it will be dependent on the flanking DNA sequences, whether or not the DNA mutated already in the region and on the DNA microenvironment. Eventually it will be like mechanics: we can always exactly determine where the planets will be relative to the earth. Likewise we will be able to exactly locate where and when the next NRM will be introduced.
Best wishes,
Peter
™ Version 4.2