Brian Miller: Co-option and Irreducible Complexity


(Bill Cole) #261

It would not and cannot. If evolution is true its mechanism has to be able to navigate and find function in exceeding large sequence spaces.

The Bill Cole Show
(John Mercer) #262

You’re still being evasive. Please stop.

(Bill Cole) #263

Were done John.

(Timothy Horton) #264

Heh. Ferrous Cranus to the bitter end. :roll_eyes:

(Neil Rickert) #265

I’m puzzled by that. I’m wondering what you mean by “trial and error”.

In ordinary usage it just means to try things out and see if they work. It does not imply a comprehensive “try all possibilities”. I use trial and error to decide what foods I like. But there are many foods that I have never tried and have no particular interest in trying.


I hear what you are saying…and you may be right…but multiplicity should give us reason to pause.

Replication…like energy from the sun…is absolutely required to lift the probability off of zero, but as long as the probability of increasing order and and the probability of decreasing order remain imbalanced, there will still be many more ways for each mutation to reduce function than ways to increase function. The only way to reduce the downward drag of multiplicity is to find a way to reduce the probability imbalance.

Again…you may be right that the wins are enough to counteract the losses {insert normal disclaimer that positive evidence exists here :-)} But if the essential imbalance is not reduced, the overall drag in fitness just has a new starting point…every mutation has a much better chance of decreasing order than it has of increasing it. The much smaller number of wins have to be really big and the much larger number of losses have to be really small to keep things in balance.

There is the added complication of the feedback happening at the level of the organism (and even entire species) as a whole and not at the individual mutation level. Even when a big win is scored, since each mutation has a much larger chance of decreasing fitness, any secondary mutations are likely to have a smaller, but still negative, long term impact on fitness. In the short term, the negative impact is overwhelmed by the win, they still contribute to a long term loss of order. In many ways the slightly negative mutations that don’t have a large immediate impact on selection are the biggest danger to the functional information content of the genome in the long term.

The degree to which the law of large numbers acting on negative multiplicity proves to be a drag on a system depends on a multitude of details that we have no real world ability assess in the world of biology…which is why the other positive evidence plays a bigger role in people’s confidence about the evolution process than an analysis of multiplicity…but it does not mean it is not playing some…or even a substantial role.

(Bill Cole) #267

Trial and error is indeed a sloppy description but lets go with it. In the case you described you are able to narrow the search by a conscious intelligent mechanism that can decide. Your search space (local restaurants and grocery stores) is also small so your chance of success (finding food you like) is high.

As the mechanism gets less directed and the search space gets lager your chance of success goes down. DMath’s discussion above appears relevant as it is the same idea that Sanford and Bessemer presented at TSZ.

(Neil Rickert) #268

But it isn’t really a search. I am combining pragmatic judgment (I like the taste) with a kind of opportunism. I don’t go around to stores to see what I should try next. But if I’m eating out and the restaurant menu offers something that piques my curiosity, then I will take that opportunity to test a new food.

And evolution is much the same – exploring those opportunities that present themselves (i.e. that are within a mutation from what is currently being used by the population). It is pragmatism + opportunism. And, in my opinion, we ought to consider that to be intelligent.

These ideas (“directed” and “search”) come from you and other evolution critics. There is no specific direction and no specific search. A population is simply exploring the boundaries of its current niche.

(Mikkel R.) #269

Excess offspring. Many more organisms are born, than can possibly survive. Through this oversampling are the rare beneficial mutations found.

You are confusing order, and possibly even complexity (in the sense of number of components) with fitness effects of mutations. You really need to separate these three concepts.

Many types of mutations with high probability are actually complexity-increasing. Most of these amount to gene duplications, or insertions, in the end, though there are several different ways of effectuating duplications. Retrotransposition, unequal crossover, strand slippage and so on can all effectively cause an increase in genetic material.

Repetitive genome segments are intrinsically more likely to yield errors in homologus recombination exactly because their repetitive nature can cause misalignment. This can cause a positive feedback loop where more and more duplications of repetitive segments pile up over time. This in turn then serves as the potential raw material for new functions to emerge, which can be beneficial to the organism.

It is noteworthy that most of the parts of the genome of different organisms thought to be so-called junk-DNA, is also highly repetitive in nature. As in, it has in large part been generated by these tendencies for repetitive DNA segments to expand due to the mechanisms of how DNA is copied.

It is thought, exactly because most mutations are actually deleterious (again, important to understand as a fitness effect, not a statement about thermodynamic “order”), that numerous sequential compensatory duplications of increasingly dysfunctional (in a relative, not absolute sense) genetic elements is also in large part responsible for the drive-up of molecular complexity seen in many cell components.

The process is called constructive neutral evolution, “CNE”. As the functions of individual genes are slowly degraded in performance because natural selection can weed out many, but can’t weed out all of the deleterious mutations, more and more copies of those components are needed to assist each other in performing these functions. Hence compensatory duplication. This has the effect of driving up the molecular and functional complexity of basic cellular components. The duplications have a positive fitness effect, because they partly compensate for the degradation in performance of point mutations happening in these genes. This also creates it’s own positive-reinforcement feedback loop. Since more and more duplications come to exist, there is now more and more genetic material that can mutate, leading to more and more degrading copies, in turn further driving up the need for more duplications to compensate.

(Bill Cole) #270

Can you model this?

Think about this exploration, the mechanism of change and the boundaries of what is changing.


That is exactly what we do with random mutations and selection.

(Timothy Horton) #272

Science has thought about and researched those things for the last 70 or so years Bill. You can read about them in any Freshman level genetics book. Well, we can read about the science. With your mental blocks you probably can’t.


The obvious answer is that it doesn’t, because the space is designed in such a way that evolution could "find’ sequences that work close by, without having to search the large space.

Does evolution have a rational explanation for that?

(S. Joshua Swamidass) #274

The laws of physics.


Why? Please explain your conclusions.

From my own understanding of biochemistry and thermodynamics, the energy from the Sun is a very important source of energy, at least in the case of eukaryotes. Prokaryotes have tricky ways of using geothermal energy, but they also need liquid water which the Sun provides. The quick version is that high energy photons produce complex and energy rich carbohydrates in photosynthetic organisms. Oxidation of those hydrocarbons drives thermodynamically unfavorable reactions, such as adding phosphate groups to adenine (i.e. AMP, ADP, ATP). These energy intermediates then drive other thermodynamically unfavorable reactions, such as the de novo synthesis of proteins and nucleotides as well as the extension of DNA molecules. If you understood how biochemistry and thermodynamics works I don’t think you would be making the claims you are making.

Thermodynamics happens naturally, so I’m not sure what you are getting at.

(Neil Rickert) #276

Oh, I have thought about it. But perhaps I don’t understand what you are intending to ask.

(Bill Cole) #277

What I am intending to talk about is the internal cellular environment that has to change in reaction to the outer environment. What allows this to change in a way for the population to react to the outer environment in a positive way.

We all agree that the structure of the animal is driven primarily by cellular DNA. How can that change to the degree that a complex adaption occurs? The population and the organism may be intelligent but how is that intelligence finding a sequence that can produce a complex adaption?

(Neil Rickert) #278

I don’t agree with that.

As far as I know, DNA is just a chemical. The structure of the animal depends on the development environment, and not just on the DNA.

(Mikkel R.) #279

If you doubt that energy creates complexity, take a pile of table sugar (which is just lumps of sucrose molecules), throw it on a frying pan and turn on the stove. Wait a bit, then come back to the black tarry substance it has turned into. The complexity will have radically increased. Where before you just had sucrose molecules, you now have large complex sheaths of asphaltene-like compounds.

You will have went from essentially having lots of these:

To having lots of this:

If you don’t think the latter is more complex than the former, then your concept of complexity is nonsensical.


In a very general sense, proteins react to the outside environment which sets off a cascade of protein to protein interactions that eventually results in protein to DNA interactions. The protein to DNA interactions cause some genes to increase their expression and others to decrease their expression. In other words, changes in phenotype in response to environmental changes are usually due to changes in gene expression.

Again, it comes down to gene expression. The network of protein, RNA, and DNA molecules and how the interact with each other is the ultimate determinant of development in complex eukaryotes. Change how those molecules interact with each through mutations and you change how development occurs.