Rapid Emergence

Loudmouth

Contributor
Aug 26, 2003
51,417
6,141
Visit site
✟98,005.00
Faith
Agnostic
First, I noted that in the case of the gorilla B could have either changed or been lost. And I could have made a mistake - I may need to add some additional variables. But, with that said, my understanding is that loss in primates may be quite high. See, for example: "Accelerated Rate of Gene Gain and Loss in Primates" by Hahn, Demuth, and Han.

What is "quite high"? 100 genes out of 30,000 genes every 10 million years? Accelerated compared to what? Unless you know what the base values are, you can't conclude much from relative terms. Also, genes only make up 3% of the genome. We do compare the other 97%.

Let's look at the differences between chimps and humans. The differences are 35 million substitutions and 5 million indels (i.e. gains/losses) for 40 million total. The genomes are 3 billion bases. 40 million of a 3 billion base genome is just 1.3%. So what are the chances that a new mutation will occur at the same place as those earlier mutations compared to elsewhere in the genome? Chances are that most new mutations will occur at bases that are the same between humans and chimps.
 
Upvote 0

whois

rational
Mar 7, 2015
2,521
119
✟3,336.00
Faith
Non-Denom
If you are going to argue that the level of "chemical complexity" in a full grown animal or plant, composed of gazillions of interconnected cells, is the same as the "chemical complexity" of a single cell or seed....



Then I'm going to have to ask you to define what you mean by "complexity"....
in my opinion an organized system of cells aren't "more complex" than a single cell, it's the fact that they somehow become differentiated during embryo development.
a group of amoebas aren't any more complex than a single one.

the word "complexity" is very hard to define.
the only real way to define it is to give some examples of what you mean, which is probably a good idea anyway.
 
Upvote 0

Resha Caner

Expert Fool
Sep 16, 2010
9,171
1,398
✟148,100.00
Faith
Lutheran
Marital Status
Married
What is "quite high"? 100 genes out of 30,000 genes every 10 million years? Accelerated compared to what? Unless you know what the base values are, you can't conclude much from relative terms. Also, genes only make up 3% of the genome. We do compare the other 97%.

Let's look at the differences between chimps and humans. The differences are 35 million substitutions and 5 million indels (i.e. gains/losses) for 40 million total. The genomes are 3 billion bases. 40 million of a 3 billion base genome is just 1.3%. So what are the chances that a new mutation will occur at the same place as those earlier mutations compared to elsewhere in the genome? Chances are that most new mutations will occur at bases that are the same between humans and chimps.

Let me back up because we may be losing the forest for the trees. If my methods are wrong, they're wrong. But I don't want that to cloud the essence of the discussion. From what little I know, it appears to me there is more than one way to assemble a tree. Would you say biology has settled on an algorithm that will only produce one tree, or is more than one outcome possible?

It seems to me there are voids in the data. I understand you see this in terms of the large amounts of data we have. I, however, don't think large amounts of data are convincing when voids in that data must be bridged with assumptions. It is those assumptions that allow for multiple possible trees. The challenge has been that it is disingenuous to just assume the trees proposed by biology are wrong - to assume an assumption is wrong - and I agree. As such, I took the mathematical models I could find (and understand) in the literature, and ran some simulations. I was honestly surprised that the probability of my alternative was as high as it was. I didn't expect it to come out that way.

You are suggesting the probability of an alternative is extremely low. In the interest of full disclosure, after my first simulation showed my alternative to be highly probable, I modified the conditions and could create scenarios where my alternative was highly unlikely. So, what would need to happen is to create a simulation that is more realistic. A lab test that simulates development of a nested hierarchy (as opposed to a mathematical simulation) would be ideal were it possible. I expect that would be extremely difficult, however. Take the gorilla, chimp, human example. Maybe the tree is correct as it is. However, inducing from one case that all trees are correct would be dodgy. I'm suggesting different conditions might achieve the same end via a different path (and demonstrated it was possible in my simulations). Until the voids are filled with data or an algorithm is found that shows only one tree is possible, I don't see how multiple paths can be ruled out.
 
Last edited:
Upvote 0

Resha Caner

Expert Fool
Sep 16, 2010
9,171
1,398
✟148,100.00
Faith
Lutheran
Marital Status
Married
... the word "complexity" is very hard to define.

It is. In my paper I defined complexity as a function of tile types. Adding a tile type increases the complexity of the system. It wouldn't be completely wrong to associate tile types with cell types. So, for each additional way in which a cell differentiates, the system would become more complex.
 
Upvote 0

SkyWriting

The Librarian
Supporter
Jan 10, 2010
37,279
8,499
Milwaukee
✟410,918.00
Country
United States
Faith
Non-Denom
Marital Status
Married
Politics
US-Others
When you define a single cell as being complex, you pretty much lost the argument.

I didn't.
It's a myth that a single cell is simple.
Especially given the example that it
holds all the information
needed for an entire organism.
 
Upvote 0

DogmaHunter

Code Monkey
Jan 26, 2014
16,757
8,531
Antwerp
✟143,395.00
Faith
Atheist
Marital Status
In Relationship
in my opinion an organized system of cells aren't "more complex" than a single cell

Then I don't know what you mean when you use the word "complex".

the word "complexity" is very hard to define.
the only real way to define it is to give some examples of what you mean, which is probably a good idea anyway.

I just did.

I consider a single cell less complex then a full grown animal build from billions of interconnected cells.

You people apparantly don't agree with that.

So, again, then I don't know what you mean by "complex".
 
Upvote 0
This site stays free and accessible to all because of donations from people like you.
Consider making a one-time or monthly donation. We appreciate your support!
- Dan Doughty and Team Christian Forums

Loudmouth

Contributor
Aug 26, 2003
51,417
6,141
Visit site
✟98,005.00
Faith
Agnostic
Let me back up because we may be losing the forest for the trees. If my methods are wrong, they're wrong. But I don't want that to cloud the essence of the discussion. From what little I know, it appears to me there is more than one way to assemble a tree. Would you say biology has settled on an algorithm that will only produce one tree, or is more than one outcome possible?

There are many different but accepted algorithms for constructing trees. While you may get different trees from different sets of data, they are never that much different for one another. You never see drastic differences where a bird and a rodent species are more closely related than two rodent species.

It seems to me there are voids in the data. I understand you see this in terms of the large amounts of data we have. I, however, don't think large amounts of data are convincing when voids in that data must be bridged with assumptions.

Could you give examples of what you are talking about?

You are suggesting the probability of an alternative is extremely low. In the interest of full disclosure, after my first simulation showed my alternative to be highly probable, I modified the conditions and could create scenarios where my alternative was highly unlikely. So, what would need to happen is to create a simulation that is more realistic. A lab test that simulates development of a nested hierarchy (as opposed to a mathematical simulation) would be ideal were it possible.

Actually, it has already been done. Different labs and companies have had their own mouse colonies for quite some time. This has allowed them to evolve and diverge. When we compare the modern mouse colonies, we find that they produce a nested hierarchy.

"Inbred mouse strains have been maintained for more than 100 years, and they are thought to be a mixture of four different mouse subspecies. Although genealogies have been established, female inbred mouse phylogenies remain unexplored. By a phylogenetic analysis of newly generated complete mitochondrial DNA sequence data in 16 strains, we show here that all common inbred strains descend from the same Mus musculus domesticus female wild ancestor, and suggest that they present a different mitochondrial evolutionary process than their wild relatives with a faster accumulation of replacement substitutions."
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1800920/

I expect that would be extremely difficult, however. Take the gorilla, chimp, human example. Maybe the tree is correct as it is. However, inducing from one case that all trees are correct would be dodgy. I'm suggesting different conditions might achieve the same end via a different path (and demonstrated it was possible in my simulations). Until the voids are filled with data or an algorithm is found that shows only one tree is possible, I don't see how multiple paths can be ruled out.

What are the specific mechanisms involved in the alternative explanation?
 
Upvote 0

SkyWriting

The Librarian
Supporter
Jan 10, 2010
37,279
8,499
Milwaukee
✟410,918.00
Country
United States
Faith
Non-Denom
Marital Status
Married
Politics
US-Others
Upvote 0

Resha Caner

Expert Fool
Sep 16, 2010
9,171
1,398
✟148,100.00
Faith
Lutheran
Marital Status
Married
Just a little follow-up ... another journal replied. Since it had been so long I wasn't expecting to hear from them, so it was nice they replied. It was also the only reply that gave me a detailed reason for not publishing, so I really appreciate that - even moreso because the reason wasn't "your conclusion is wrong" but "your conclusion is too obvious."

As I said, I was trying to keep the first step simple, but apparently I made it too simple. In the reply, the reviewer noted I could have used a Markov chain model to evaluate the same question instead of using aTAM, and the Markov model would have given me an exact (closed-form) answer rather than my empirical answer. In fact, the reviewer gave that exact answer and showed it was the same as my emprical answer. For me, it was really cool just to know the idea worked.

On the flip side it also confirmed what I had sort of suspected - that my mathematical knowledge is insufficient to take this any further. Still, I wonder how willing the reviewer would be to chat with me - given I would basically just be wasting his/her time in order to satisfy my own curiosity. Given there is a closed-form answer to the problem I posed, I wonder if the asymptotic relationship I suggested exists. And if so, what does that mean? But ... since it was a blind review the journal would have to agree to give me the reviewer's contact info. That seems unlikely to me.
 
Upvote 0