Convergent Evolution - Reasonable conclusion? or convenient excuse?

Hi all,

I can pretty easily write a program to generate genomes for the PARS program to analyze, but I need feedback on a few things. This list is for producing a final population through normal reproductive descent:

1. I'll go with the standard 4 nucleotides: CAGT

2. Genome size as measured by number of nucleotides will be large. This will be settable per program run.

3. Mutations will be restricted to single nucleotide substitutions, with no additions or deletions.

4. Number of mutations per offspring will be random with a settable upper and lower bound.

5. Number of offspring produced by a single organism per generation would be random with a settable upper and lower bound.

6. Position of mutations will be random.

7. Number of generations will be settable.

What would be a good genome size in nucleotides for the PARS program to analyze? What would be a good number of generations? What would be a good number of offspring produced per organism per generation? What would be a good final population size?

For the designer case I'll start with a certain number of individuals, then take them through enough generations to give the same size final population as in the evolution case. I could use Bob's feedback on this. How many original individuals should I start with? How many nucleotide sequences should they have in common?

The performance of the PARS program is a consideration. If I ended up with final population sizes of a billion individuals with genome sizes of a billion nucleotides I assume the PARS program would never finish. What would be a reasonable genome size and a reasonable final population size? Now that I think about this, I think I'll need to have random death and a fixed maximum population size. If all individuals always reproduce and all offspring always live the tree would grow too big.

And most importantly, predictions. What do we predict we should see from the PARS program for both cases? In other words, if I handed you sets of genomes of the individuals of final populations, one set each for both standard evolution and design, would you be able to tell which was which from the PARS analysis? I'm challenging you (or WK or someone) to develop the criteria for telling the difference. Then I will provide you the genomes for 100 final populations. Half will be evolution, half will be design, and you have to figure out which is which. If your criteria are unable to distinguish then you'll get roughly 50% correct. If your criteria are perfect then you'll get 100% correct. What do you think is a reasonable criteria that would demonstrate that the PARS program can tell the difference between populations that developed solely through descent versus from a set of originally designed individuals.

--Percy

Mr. Jack writes: In real data, there'd be many more datapoints, and more states; usually allowing multiple changes in the same character to the be resolved.

True. It doesn't model evolution at all. Each individual mutates randomly on one seventh of its genome in only 6 possible ways, and there's no selection, so no conservation. Run it on until there are a million individuals, and it would hit some individuals with the exact same genomes as the original 15 depicted, and other chance genetic "twins".

Also, relationship to an ancestor more than seven generations back will probably be undetectable.

Sooner or later, you get an organism AAAAAAA that bears no resemblance to ancestors ten generations back, but is a replica of the original!

Using all 26 letters of the alphabet would seem to work much better, but organisms would still eventually change out of recognition to their ancestors if there was no conservative selection of traits going on.

Perhaps random genomes of any size could be fit into nested hierarchies by a good program.

Yes, they can.

The evolutionary prediction goes beyond just saying that we can make a nested hierarchy. We might summarize it as follows:

* If we construct two trees using two different largish randomly-chosen data sets the trees will be in good agreement.

* The larger the data sets, the closer the agreement.

* This will also agree with the trees that we would sketch out if we did so only by considering the intermediate forms in the fossil record (where available, for example in vertebrates).

You might like to read my article on taxonomy, especially the section on robustness.

Thanks Wounded King, Mr. Jack, Percy

Adding in the ancestral sequence as an outgroup still doesn't give me Percy's tree exactly, but it does produce a tree with a length of 14.

And Mr. Jack had one with a tree length of 12 (Message 81).

I had wondered if making an outgroup would help, and if adding a lot more 'a's to the lineages (to more completely model the relatively small proportion of the genome that mutates).

Mr. Jack Message 90: The outgroup is important because it helps the program determine which states are later mutations and which states were there originally. The second reason is that Percy's example happens to have multiple mutations at the same site, specifically the first mutation on the left branch is then overwritten by the second mutation on the leftmost branch (AAAAAAA->AAAAABA->AAAAAGA).

Which would be a rare occurrence in a real world example, so we can avoid this in making future models (or use a longer genome and a random generator for location and change).

Percy Message 91: I can pretty easily write a program to generate genomes for the PARS program to analyze, but I need feedback on a few things. This list is for producing a final population through normal reproductive descent: I'll go with the standard 4 nucleotides: CAGT Genome size as measured by number of nucleotides will be large. This will be settable per program run. Mutations will be restricted to single nucleotide substitutions, with no additions or deletions. Number of mutations per offspring will be random with a settable upper and lower bound. Number of offspring produced by a single organism per generation would be random with a settable upper and lower bound. Position of mutations will be random. Number of generations will be settable. What would be a good genome size in nucleotides for the PARS program to analyze? What would be a good number of generations? What would be a good number of offspring produced per organism per generation? What would be a good final population size? For the designer case I'll start with a certain number of individuals, then take them through enough generations to give the same size final population as in the evolution case. I could use Bob's feedback on this. How many original individuals should I start with? How many nucleotide sequences should they have in common?

Given that this thread is about convergent evolution, perhaps we need to start with two different genomes and run the mutations/generations, with come mechanism to select for a specific output?

For the designer model we would also need to see some predictions of what the design would cause:

1. start with a set of different genomes
   
2. copy of a mutation from one genome to another to model reuse of design
   
3. fewer generations should be needed for the designer model

Perhaps work backwards from the final results of the evolution model, so we have the same end but different hypothetical beginnings.

Enjoy.

---

we are limited in our ability to understand
by our ability to understand
Rebel American Zen Deist
... to learn ... to think ... to live ... to laugh ...
to share.

---

• • • Join the effort to solve medical problems, AIDS/HIV, Cancer and more with Team EvC! (click) • • •

And Mr. Jack had one with a tree length of 12 (Message 81).

Note that the tree Wounded King is referring to is the one in which he added the outgroup, whereas I was referring to the one without.

Testing ...

Presently when I click on "All Topics", I just get the World Cup thread, that being the last thread I commented on / the last thread anyone commented on.

I'm just posting here to see if I get the same result with this thread before I contact Percy.

This has nothing to do with convergent evolution, do carry on.

quote:

---

I don't think convergence (or lack thereof) constitutes evidence for or against ID. It would constitute evidence against darwinism if found in significant quantity.

---

And why would that be evidence against 'darwinism'. Please define 'significant quantity'.

Hi ramoss, good question, hope you get an answer.

And why would that be evidence against 'darwinism'.

Because he said so. Of course this would be 'darwinism' according to his interpretation, and not necessarily evolution as used in science.

Please define 'significant quantity'.

A lot more than we have seen:

quote:

---

My Message 21:
http://www.guardian.co.uk/...sep/29/number-of-living-species

quote:

---

The number of species on the planet that have been documented by scientists has risen to 1.9 million, according to the world's most comprehensive catalogue of plants and animals. ...

...
While there are only 100 instances listed in the wiki article, and I don't call 100/1.9X10^{^6} = 0.005% a significant problem.

---

BobTHJ has not really answered my previous post on this issue, but has gone off on other tangents.

And I would say that the (documented) extremely LOW incidence of convergent evolution examples is evidence against the IDologist claims of "reuse of design" - especially when there are numerous examples of opportunities for "reuse of design" to have been used that have been missed.

Enjoy.

---

we are limited in our ability to understand
by our ability to understand
Rebel American Zen Deist
... to learn ... to think ... to live ... to laugh ...
to share.

---

• • • Join the effort to solve medical problems, AIDS/HIV, Cancer and more with Team EvC! (click) • • •

ramoss writes: BobTHJ writes: I don't think convergence (or lack thereof) constitutes evidence for or against ID. It would constitute evidence against darwinism if found in significant quantity. And why would that be evidence against 'darwinism'. Please define 'significant quantity'.

Does anyone remember about 2 years ago when creationist/I.D.ist Randman spent an entire thread arguing that there wasn't enough convergent evolution for "Darwinism"?

That argument went along the lines of "if it's possible for invertebrates to evolve into vertebrates once long ago, then they should've done it again by now".

We should get Randman and Bob together so they can come up with a grand unified creationist theory of why naturalistic evolution didn't happen.

"There is both too much and too little convergence for "Darwinism" to be right, doubly proving it wrong".

I wouldn't use natural selection to describe the vast differences across species especially when it lists the same type of animal that is classified as such based almost identical features in appearance and then state they are unrelated.

This list does not make any sense whatsoever and if it is suppose to be presented as the facts then good luck on convincing anyone of this crap.

I wouldn't use natural selection to describe the vast differences across species especially when it lists the same type of animal that is classified as such based almost identical features in appearance and then state they are unrelated.

But they are unrelated. DNA doesn't lie.

Hi barbara, and welcome to the fray.

I wouldn't use natural selection to describe the vast differences across species especially when it lists the same type of animal that is classified as such based almost identical features in appearance and then state they are unrelated.

Not sure I understand you here, could you expand your argument some, and provide reference to an example from my post?

This list does not make any sense whatsoever and if it is suppose to be presented as the facts then good luck on convincing anyone of this crap.

Well that's one way to deal with not understanding the issue.

It's just a list of instances where some similarities have evolved in different species, some of them with a high degree of similar appearance.

Again, I'm not sure what your issue is, so you may want to expand on this as well.

Enjoy.

... as you are new here, some posting tips: type [qs]quotes are easy[/qs] and it becomes: quotes are easy or type [quote]quotes are easy[/quote] and it becomes: quote: quotes are easy also check out (help) links on any formatting questions when in the reply window. For other formatting tips see Posting Tips If you use the message reply buttons (there's one at the bottom right of each message) your message is linked to the one you are replying to (adds clarity). You can also look at the way a post is formatted with the "peek" button next to it.

---

we are limited in our ability to understand
by our ability to understand
Rebel American Zen Deist
... to learn ... to think ... to live ... to laugh ...
to share.

---

• • • Join the effort to solve medical problems, AIDS/HIV, Cancer and more with Team EvC! (click) • • •

I am confused I thought DNA analysis cannot specify a date of when any mutation occurred. If genes were deleted and replaced with new mutations, how do you they ever existed?

Quite often the answer is like old copy machines. The machine starts and instead of making just one copy it just keeps running, cranking out copy after copy of the same page.

That happens in the real world as well, an error happens and two or five or ten copies get made. Some of those excess copies also have errors and so even though the original function is maintained the extra copies can introduce something entirely new.

---

Anyone so limited that they can only spell a word one way is severely handicapped!

I can't really follow your reasoning here, Taq wasn't saying that he could identify any specific date that a particular mutation ocurred, simply that there is no reason to presuppose that all of the mutations needed to occur simultaneously in the same organism. What can be done is to identify the most parsimonious model for the diversification of related sequences from an ancestral sequence.

The way deletions are identified in genomic analysis, when they do occur, is by comparison to reference sequences. These will usually be taken from a consensus sequence from a sample of the population or in cross species comparisons from a putative conserved ancestral sequence.

In most cases we also have to infer the ancestral sequences from extant sequences, although now there are some long term data sets, the Lenski long term evolution experiment one in particular, where we do know exact ancestral sequences.

If genes were deleted and replaced with new mutations

That isn't what happens, as Jar suggests the mutations are alterations of existing genes, not the deletion and wholesale replacement of one gene by a totally novel gene.

TTFN,

WK