Evolution: time for some change?

I mainly stole the phraseology of "meta" to describe an element of functional redundancy across a range of species, with potentially different molecular characteristics.

When I referred to networks I meant the regulatory links between factors (such as the protein from gene A activates transcription of gene B or the protein from gene X activates the protein from gene Y by phosphorylation).

The expression patterns generated by a network would therefore be a function of the regulatory interactions and the initial state of the system (including the distribution of various cell types expressing various different molecules).

I didn't intend to imply any experimental similarity between the Stuart paper and what I was suggesting. I meant mapping the functional interactions/regulatory links through a combination of misexpression, genetics, and biochemistry. All these are necessary to determine the various effects of molecules and more importantly the direct interactions responsible for those effects.

For instance removing a transcriptional repressor P can lead to LOSS of expression of gene R that is not even a target of that molecule. This could make one think P was an activator of R, when it is neither an activator or is R a target. Alternatively P could have been repressing Q, a direct repressor of R. Upon loss of P, Q is no longer repressed and therefore is expressed and can repress R.

Therefore one has to hunt down the molecular details of regulatory effects beyond a genetic description on effects upon expression patterns (although w/double mutants and detailed staging one can figure much out, but that's much more tricky and molecular handles make such inferences more straightforward to demonstrate).

 

Log in or Sign up to hide all adverts.

I read your friend Isaac's paper. I can see why no one voluntarily reads his stuff. His writing might be even less clear than mine.

It was interesting, it'd be nice if they measured some of the parameters they picked and showed they were in the right range. Still I think the work goes in the right direction.

I hope more for molecular insight into the actual mutations that lead to change, but that is certainly asking a lot. It would be really cool if they could demonstrate molecular compensation between different factors such as SHH and FGFs between mouse and vole or something along those lines.

The other concern with modeling is related to your comment on how there are many different ways to achieve a certain result. Just as molecular details can change and maintain the same macroscopic pattern, models differing from the true details can yield results consistent with what is known from experimental work. Although certain approaches may work, they may never be used in evolution because they are overly sensitive to parameter choice and therefore are difficult to establish.

 

Log in or Sign up to hide all adverts.

some inane ramblings

i knew this would be a productive discussion.

Congrats on your successful proposal defense, scilosopher.


Re: microarray analyses and discovery of co-evolving genes (I am assuming that co-expressed genes will also be coevolving although I'm not sure of the validity of this assumption yet);
I'm pretty sure there are many people who are using this technique to identify, not only gene pairs that are co-expressed, but also genes involved in entire metabolic pathways, including their regulation? I would think that these genes would evolve as a unit and not merely as individual genes. I therefore think that a modular view of evolution for these genes is an excellent research approach.
co-expression, in itself, may indicate functional linkage, but there are many situations in which genes that are not functionally linked would co-express. However, that Stuart et al. found this coexpression across multiple phyla, suggests strongly that these genes are functionally linked, which I find to be an incredibly important discovery.

Re: the gene conservation across evolutionary time;
Sci-Phi is right, important genes are conserved through natural selection. I think structural protection of DNA from mutation, in eukaryotes, like methylation, are somatic cell adaptations to protect those genes or regulatory sites within the life and cell proliferation of the organism.

To get back to the original topic; although I agree that 'change in gene frequency' is not a fully descriptive definition of evolution, it is at least fully accurate.

 

Log in or Sign up to hide all adverts.

Re: some inane ramblings

the definition 'change in morphology' is also fully accurate.

 

Wouldn't preadaptive changes (someone mentioned them just recently I think) represent a change in gene frequency (or allele frequency or whatever) without necessarily representing a morphological change?

 

I think the morphology view is hard to argue for. There is the potential for many functional changes that could impact fitness that wouldn't effect morphology. For example alterations in metabolic pathways have no effect on morphology, but would affect fitness and should certainly be considered in evolution.

By the same token it isn't clear that gene frequency is completely satisfactory either. Alterations in arrangement of genes could facilitate downstream changes that are of large functional significance. Although this would presumably have an impact on allele frequency, it is the first point of change with functional consequence.

Further there are some clear examples of epigenetic changes that have hereditable consequences. Most of those studied constitute switches that cycle between two states, but it would be relatively difficult to study one time epigenetic mutations (of course so far as I know Jacob-Krutzfeld and prion diseases in general are a one way event).

Pragmatically the ability to isolate genes by PCR and the fact that DNA sequence can be read as a discrete measure of type for any population makes it a very good choice for an evolutionary metric. By the same token evolutionary studies in the absence of understanding the functional impact of the changes noted are of only academic interest.

I think it's good to encourage studying evolution in the context of change in morphology or system function, but I do not see any other measure that has nearly as many pragmatic advantages as something DNA based. By the same token with increased genome sequencing, I think other types of changes are worth study (ie arrangement, repeat sizes and distributions, cooccurance of allele pairs, and the like which all can be done with PCR on a population level ... and may be for all I know).

(thanks paulsamuel)

 

Monkey, do you mean morphology with respect to protein conformation, or with respect to the phenotypic expression of the entire organism?

 

Re: Re: some inane ramblings

not necessarily. a population can have gene frequency changes with no consequent morphological changes. This is evolution.

 

a population can also have genotypic change without phenotypic change. This is evolution.

I'm merely pointing out the emptyness of the statement that the change in gene frequency is fully accurate. Lots of things are fully accurate, but not necessarily satisfactory. Morphology is a fully accurate description of evolution. In a way it is more accurate than the genotypic change because we actually have some historic record even of the morphological change, although not complete (fossil record). The historic record for genotypic change is very sparse.

so let us not bullshit around with clinging to straws. We need to examine the core of evolution.

 

You can certainly argue that various metrics of evolution have their own advantages. Like morphology and the fossil record or DNA based changes which I mentioned some advantages of in my last post.

Our working conception of evolution must be oriented towards metrics in order to do science. That's a pragmatic necessity. Our theoretical conception is more free of this constraint, but is somewhat academic until one can experimentally address the theoretical subtleties that lie outside our working conception as defined by various experiments.

Ideally evolution in my mind should seek to determine the primary causal material (ie within the organisms "design") and contextual (ie within the environment) changes that alter the fitness of individuals within a poplation. We obviously have a very poor grasp of environment in many cases simply due to the complexity of factors that are important. We also have sever limitations on what measurements we can make.

Often in science, as in this case, the important thing to do is see what straws you can get your hands on and make intelligent use of them. By the same token if the availability goes from straws to hay, so to speak, one is an idiot to stick to looking at things only at the level of the straws as you're accustomed. It's equally stupid to ignore what you know from your past work and ignore what you've learned from looking at straws or try to take on too much.

It's clear that morphology can sometimes be deceptive without using molecular markers and examining mechanism, even when dealing with spatial aspects of evolution. The important thing to do is develop integrative approaches, not say "change in allelle frequency is fully accurate" or "change in morphology is fully accurate".

The only thing that's a fully accurate depiction of evolution is tracing the path of every atom since the beginning of life. We can't do that. Things have to be coarse grained at a certain level, when you're dealing with something as complex as biology (and possibly the various levels studied hierarchically and mentally integrated). I personally think you shouldn't say A is better than B, rather A can guide us in this way and B in that and the best way to proceed is get information X from A and then fill in Y from B.

It's also a good way to avoid wasting time arguing ... put forward what is learned from what and then it will become clear that certain information is more useful in understanding various different questions.

 

spuriousmonkey said:

But evolution is only a description of a series of relatively unrelated processes, it's more of a pattern than it is a force. A multidisciplinary approach may be the only hope in studying it completely. Genetics and morphology, as you say, aren't as... uh... bijective(?) as they could be in our present study of them. Given that we aren't able to relate them in a one-to-one manner, a study of evolution would seem to require both.

 

I agree that the definition 'gene frequency' is not fully descriptive, and thus not fully satisfactory, but I cannot agree that morphology is an acurate descriptor because there are evolutionary changes that are not manifested in phenotype.

Also, one of the reasons why DNA sequence phylogenetics are so powerful is because of the paucity of the fossil record (evolutionary relationships can be examined with NO fossil record). So, it seems to me, your argument, that morphology is the more accurate description because of the fossil record, is contrarian (i.e. it actually argues the opposite point, that morphology cannot be the more accurate description because of the incompleteness of the fossil record).

 

studies of evolution do use both, but rarely simultaneously. comparative morphology is a powerful tool in studying evolution, and it was the only tool evolutionary biologists had before the 1960's.

 

i didn't say that.

 

evolutionary relationship is not the same as evolutionary history. You can also make a nice evolutionary relationship scheme based on morphology.

but we are still digressing. I shall repeat myself.

hence let us not get overy excited by the 'accurate nature' of change in gene frequncy. We might be able to measure it (similarly we can measure morphological traits rather accurately) rather accurately, but that doesn't mean that we know what it means.

 

I don't think lots of things are fully accurate (except semantically depending on various definitions). Every measure and explanation is approximate and rough. The details are way beyond our experimental and theoretical abilities right now.

The question is what would you find a satisfactory measure for understanding evolution?

Personally I think people should take well studied networks in a given organism and study them in well separated organisms. Try to bootstrap themselves as much as possible off what is known in the original organism. For an interesting paper that takes that approach see:

Regulation of the Tribolium homologues of caudal and hunchback in Drosophila: evidence for maternal gradient systems in a short germ embryo.

Wolff C, Schroder R, Schulz C, Tautz D, Klingler M.

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=9716530&dopt=Abstract

 

It seems that evolution discussions may restart Lamarck-Darwin battle, eh? I'm doubting that natural selection is the only way of evolution, considering that there are now just plain too many specieses (or should I say: distinct speciese?) for natural selection to be true..... Help.

 

Did Lamarck posit a higher rate of development/speciation? Haven't read his stuff...

 

Natural selection isn't supposed to explain how you get new variants. I haven't read Darwin's stuff myself (started once and never quite got through it), but my understanding is that he gave no force for generating variants to be selected upon, and rather presumed their existence.

So the number of species that exist is something apart from Darwin's theory of natural selection. I don't think Lamarck is a rational alternative. Genetics has lead to the understanding that variants are generated by imprecise maintanance and copying of existing DNA as well as its "shuffling" in the recombination process during sperm and egg production.

To get evolution one needs a balance of variants and consistency. Too inconsistent and a species doesn't maintain it's past abilities, too consistent and it doesn't have variants that can handle new or changing conditions.

 

Lamarck, as I recall, also doesn't have a very good explanation for sympatric speciation, that is, why two groups of creatures that are originally in the same population will evolve to be different enough from one another that they can no longer interbreed... Lamarck's ideas, as I recall, only related to positive pressures on the population. I could be wrong though.