Non-mendelian genetics/ non-darwinian evolution

Hi WK,Just a thought or two about epigenetics, a concept I have only recently come across, but find quite intriguing (so no doubt you and Mammuthus can put me straight on a couple of things ).

quote:

---

I was thinking along the lines of epigenetic effects such as DNA methylation, which is not only responsible for your imprinting but for other heritable changes, or heritable histone alterations of chromosome structure. Many of these epigenetic factors may not be heritable in the long term, at the moment it is not clear to what extent epigenetic factors are generally heritable or how stable those factors are.

---

From the little reading I have done on the subject, it occurs to me that although the idea that epigentic factors can have a direct effect on phenotype is non-Mendalian (not to mention non-Darwinian) at first glance, the ability to produce offspring which could vary according to environmental changes could be a selectable trait in itself. In environments where there is a high amount of change it would be advantageous to be flexible in the kind of offspring you produce (very similar to the way that sex ratios change in reptiles and birds). On the other hand if a specific morphology (for example) was trumping all others then it would make sense to 'fix' it in place with a more permanent mutation. In a review I read a while back (which I now can't find ) it suggested that this variability followed by fixation could be a possible 'fast-track' to speciation.Any thoughts?

Hello Again,Thanks for the links, and for elaborating on what I was trying to say: I tend to over-simplify sometimes. For example:

quote:

---

To me you seem to be discussing something along the lines of 'evolvability'

---

...now that was the word I was looking for! Having said that, I'd like to make an observation about your definitions of epigenetics.I think the reason I like to lump them all together (or at least draw a fuzzy grey line between the two), is that they could share common mechanisms and that those mechanisms (although heritable) are essentially reversable - DNA can get methylated and demethylated, histones can be deacetylated and acetylated.If you imagine a situation where environment caused imprinting to occur, and that imprinting was only partially inherited then this gives a possible source of phenotypic plasticity. Not really backed up by any evidence as such, but maybe a possibility?

Sorry about not replying to your earlier post sooner, my typing tends to be as speedy as a sloth on tranqs, so I don't tend to post very much I think this is quite a useful definition to work with. If the amount of non-sequence modification doesn't matter (as in proper 'junk' DNA sections), then it doesn't have to be thought about in relation to evolution/inheritance.If the amount of modification is vital for the gene to function properly then it can be treated for all intents and purposes like a Mendalian allele (but maybe with a little more potential to 'fine-tune' a phenotype). Whether it compliments or antagonises the traditional sequence 'genotype' or not doesn't matter - it is at least 'pseudo-Mendalian'. I think I'm with PS on this one - trying to assign traits to a genome where genes are interacting like crazy is flipping complicated.However, if there are a number of methlylation states which can all be influenced by the environment, and can all give rise to viable (yet differing) phenotypes then the waters can get a bit murky IMO. I agree with an earlier comment of yours: hopefully the Evo-Devo side of things can start to clear things up a bit.Edits: an attempt to make sense This message has been edited by Ooook!, 10-26-2004 11:31 AMThis message has been edited by Ooook!, 10-26-2004 12:08 PM