Sal's Flower?

[T_aquaticus]

@colewd

Not against misrepresentations and logical fallacies as @T_aquaticus has demonstrated above.

If you can understand the Birthday problem you will start to understand why gpuccio is wrong and why “coordinated mutations” is so misguided.

Birthday problem

[John_Harshman]

@colewd

Their are reason for DNA the Introns and the Spliceosome. Its to make a wide range of proteins to allow for life diversity. If almost all protein sequences were equally functional this complex machinery would not be necessary.

Seriously: what are you trying to say? You can’t explain word salad by slightly longer word salad. And we seem to be getting farther and farther from the topic of “Sal’s flower”.

@colewd

I am not going into the blast math at this point but to get an idea to translate to bits take a human protein with 100 amino acids and a mouse with the same positions except for 10. The number of bits of human functional information would be 20^90 or about 2^390 or 390 bits.

So it’s just a bizarre and non-intuitive way to describe sequence similarity. How does this differ in any meaningful way from just saying “90% similar”? How is that, in any meaningful sense, information? And note that calling it information assumes that we’re randomly choosing from among all 20 amino acids at each site.

[T_aquaticus]

@colewd

The issue (as @gpuccio describes) is not the size of the Island. It’s the ratio of the size of the island to the ocean. The Island being relevant function and the ocean being all possible combinations.

The problem is that @gpuccio is blind to the number of islands in the ocean. This is the Sharpshooter fallacy.

[Rumraket (Mikkel R.)]

@colewd

The issue (as @gpuccio describes) is not the size of the Island. It’s the ratio of the size of the island to the ocean.

The difficulty of understanding this intuitively is that numbers get much larger than things we can imagine like Islands and oceans.

And thus you reveal that you have taken it as an axiom that this ratio is extremely large, and that this particular island is a target in a sea of nonfunctionality, as opposed to just one among many different islands evolution happened to sample.

Texas
Sharpshooter
Fallacy

[John_Harshman]

@T_aquaticus

The problem is that @gpuccio is blind to the number of islands in the ocean.

Be fair. He’s also blind to their sizes and their connections.

[Roy (30-year veteran)]

@colewd

I am not going into the blast math at this point but to get an idea to translate to bits take a human protein with 100 amino acids and a mouse with the same positions except for 10. The number of bits of human functional information would be 20^90 or about 2^390 or 390 bits.

That would mean the amount of human functional information in a protein depends on which human you’re comparing it to - and can be both 20^90 and 20^91 and 20^89 and other values all at the same time.

It also means that some chimp proteins contain more human functional information than the equivalent protein in some humans.

It’s a completely useless number whose only purpose is to look ‘sciency’.

Out for the day.

Out for the count, more like.

[Chad_the_Layman (Chad Stewart)]

@Rumraket

To ID proponents these statements are effectively interchangeable, since you’re all laboring under the delusional axiom that functional islands are so small and far apart they couldn’t possibly be hit by blindly sampling sequences in that space.
Since the functions are nevertheless found (as in, they exist in life and can be shown to be conserved in certain clades on phylogenies) you guys say we must infer that someone must have been “aiming” for them.

By focusing the argument specifically on a protein like ubiquitin and it’s variants (rather than anything that works ), you have drawn a target around it, and now argue someone must have been aiming because your “target” was hit. Because with your unsupported axiom that the “targets” are extremely small and rare, they couldn’t have been hit by just blindly shooting into sequence space.

That’s the ID argument in a nutshell, and it’s why it really does commit the Texas Sharpshooter Fallacy. As soon as you pick a concrete example (a known protein sequence including all it’s variants used to extrapolate a rough size of it’s “island”) with some function and then try to extrapolate the prior probability of hitting that “target” in sequence space with blind shots without aiming, you are committing the fallacy . Doing that is what commits the fallacy.

It’s like drawing a particular set of cards from the deck, finding that it’s a useful set of cards in your particular situation, then calculating the odds of that draw after having drawn it, then saying the odds of that draw is so small the deck must have been stacked on purpose to yield that set of cards. This would be analogous to what you are doing with sets of cards (protein sequences).

Responding to say LOVE THIS because I can only press the heart button once.

[colewd (Bill Cole)]

@John_Harshman

Seriously: what are you trying to say? You can’t explain word salad by slightly longer word salad. And we seem to be getting farther and farther from the topic of “Sal’s flower”.

What I am saying is the variation we are seeing is designed in the system. Both alternative splicing and transcription and translation create protein variation which in turn creates morphological variation. Sal’s flower shows lots of genetic variation between humans, mice, chickens and zebra fish.

You are claiming the variation comes from various forms of random change where I am claiming the variation comes from different starting points. Your challenge is to show there is enough variation in reproduction to form the observed pattern. The biggest challenge is accounting for genes being added given genes code for proteins that count on a functional arrangement of amino acids.

@John_Harshman

So it’s just a bizarre and non-intuitive way to describe sequence similarity. How does this differ in any meaningful way from just saying “90% similar”? How is that, in any meaningful sense, information? And note that calling it information assumes that we’re randomly choosing from among all 20 amino acids at each site.

Bits are standard measurement for information. In this case we are talking about functional information.

Gpuccio’s method assumes a similar starting point for the genes he analyzes. In many cases we see a lack of variation in those genes that he attributes to functional constraint. He measures this with a Ka/Ks ratio where we see many more fixed neutral mutations.

Functional constraint indicates a higher level of functional information.

[John_Harshman]

@colewd

What I am saying is the variation we are seeing is designed in the system. Both alternative splicing and transcription and translation create protein variation which in turn creates morphological variation. Sal’s flower shows lots of genetic variation between humans, mice, chickens and zebra fish.

More word salad. Transcription and translation do not create protein variation, though alternative splicing does, to the limited extent that it actually happens. None of that has anything to do with Sal’s flower.

@colewd

You are claiming the variation comes from various forms of random change where I am claiming the variation comes from different starting points.

Once again you are confusing the source of variation with the pattern of variation. The pattern supports common descent, and I am supposing that by “different starting points” you mean separate creation. Separate creation can’t account for the pattern of distribution on Sal’s flower.

@colewd

Your challenge is to show there is enough variation in reproduction to form the observed pattern.

Pseudogenes are good evidence for that, aren’t they? They’re all evidence of gene loss.

@colewd

The biggest challenge is accounting for genes being added given genes code for proteins that count on a functional arrangement of amino acids.

There’s an extensive literature on the evolution of new genes. Have you ever cracked any of it open for a look?

@colewd

Bits are standard measurement for information. In this case we are talking about functional information .

Why should a measure of similarity be interpreted as a measure of functional information or as a measure of any sort of information?

@colewd

He measures this with a Ka/Ks ratio where we see many more fixed neutral mutations.

Note, incidentally, that all this assumes common descent. Why don’t you reject it as mere speculation?

[colewd (Bill Cole)]

@John_Harshman

Separate creation can’t account for the pattern of distribution on Sal’s flower.

Your argument appears to be hanging on this assertion. Can you tell me why you don’t believe that the pattern in Sal’s flower can be the result of different starting points?

[T_aquaticus]

@colewd

Can you tell me why you don’t believe that the pattern in Sal’s flower can be the result of different starting points?

Parsimony. The pattern is consistent with common ancestry and evolution.

For, be it observed, the exception in limine to the evidence which we are about to consider, does not question that natural selection may not be able to do all that Mr. Darwin ascribes to it: it merely objects to his interpretation of the facts, because it maintains that these facts might equally well be ascribed to intelligent design. And so undoubtedly they might, if we were all childish enough to rush into a supernatural explanation whenever a natural explanation is found sufficient to account for the facts. Once admit the glaringly illogical principle that we may assume the operation of higher causes where the operation of lower ones is sufficient to explain the observed phenomena, and all our science and all our philosophy are scattered to the winds. For the law of logic which Sir William Hamilton called the law of parsimony—or the law which forbids us to assume the operation of higher causes when lower ones are found sufficient to explain the observed effects—this law constitutes the only logical barrier between science and superstition. For it is manifest that it is always possible to give a hypothetical explanation of any phenomenon whatever, by referring it immediately to the intelligence of some supernatural agent; so that the only difference between the logic of science and the logic of superstition consists in science recognising a validity in the law of parsimony which superstition disregards. Therefore I have no hesitation in saying that this way of looking at the evidence in favour of natural selection is not a scientific or a reasonable way of looking at it, but a purely superstitious way.
–George Romanes, “The Scientific Evidences of Organic Evolution”, 1882

Your argument was already dealt with 140 years ago.

[John_Harshman]

@colewd

Can you tell me why you don’t believe that the pattern in Sal’s flower can be the result of different starting points?

Yes, but haven’t I told you many times already? I’m assuming that by “different starting points” you mean separate creation of kinds. Is that right? If so, then we do not expect a nested hierarchical pattern from separate creation, but we do expect one from common descent. What part of that is unclear? Why do you always ignore that point?

[Roy (30-year veteran)]

@colewd

You are claiming the variation comes from various forms of random change where I am claiming the variation comes from different starting points. Your challenge is to show there is enough variation in reproduction to form the observed pattern.

Then your challenge would be to show there isn’t. But somehow it’s only other people who have to provide support for their views.

[colewd (Bill Cole)]

@John_Harshman

If so, then we do not expect a nested hierarchical pattern from separate creation,

Instead of labeling the pattern a “nested hierarchy” which is invoking a labeling fallacy can you argue why special creation could not create the pattern we are observing.

This appears all based on methodological naturalism as @T_aquaticus is arguing?

Is it possible that scientific inquiry is validating Genesis 1’s claim of created kinds once you remove the restriction of methodological naturalism to the investigation?

[Mercer (John Mercer)]

@colewd

Functional constraint indicates a higher level of functional information.

That’s objectively false, as I previously showed you with human beta-cardiac myosin vs. actin.

Neither you nor @gpuccio has addressed that.

[T_aquaticus]

@colewd

This appears all based on methodological naturalism as @T_aquaticus is arguing?

It’s based on parsimony. Read the quote again.

[John_Harshman]

@colewd

Instead of labeling the pattern a “nested hierarchy” which is invoking a labeling fallacy can you argue why special creation could not create the pattern we are observing.

First, this is not a labeling fallacy. Nested hierarchy has a clear definition, and the data clearly fit that definition. Notice how I can account for almost every feature of the distribution by a single change on the standard tree. That’s not chance and it’s not just labeling. There is a real pattern here, and that pattern has a name.

Second, of course special creation could create the pattern; special creation could create any pattern at all. It’s just that, in contrast to common descent, we have no reason to expect any particular pattern, and certainly not a very strong instance of the one precisely predicted by common descent. Special creation has no expectations and can explain nothing. It could still be true anyway, but we have no reason to think it is.

@colewd

This appears all based on methodological naturalism as @T_aquaticus is arguing?

I don’t think he’s arguing any such thing, and it isn’t. Except to the degree that the naked hypothesis of creation can’t be tested at all, since it makes no predictions and cannot engage with data. If methodological naturalism is merely another way of saying that only hypotheses capable of being tested by data should be entertained, then yes, this is based on methodological naturalism. But if we start accepting untestable hypotheses, we have abandoned any sort of science entirely. Is that really what you want?

@colewd

Is it possible that scientific inquiry is validating Genesis 1’s claim of created kinds once you remove the restriction of methodological naturalism to the investigation?

No, it’s not possible. The pattern is what we expect from common descent. Interpreting it as creation would seem to require intentional deception on the part of the creator.

[Rumraket (Mikkel R.)]

@colewd

Instead of labeling the pattern a “nested hierarchy” which is invoking a labeling fallacy

Bill are you okay? This is unusually incoherent even for you. No, calling a pattern by it’s proper name is not a “labeling fallacy”. Had @John_Harshman called a sinus curve that, that would not constitute a fallacy. The proper name for a nested hierarchy is a nested hierarchy. So when referring to such a pattern, one should simply use that word.

@colewd

Is it possible that scientific inquiry is validating Genesis 1’s claim of created kinds once you remove the restriction of methodological naturalism to the investigation?

Yes if we toss away the restriction against untestable ad-hoc reasoning, everything follows, not only Genesis 1. Also Odin, Zeus, fairies, witches and occult powers, timetravelers, hyperadvanced alien species that create simulations for us to live in, the Pantheon of the God’s of Chaos, the wizard Merlin, noted fallen space marine apothecary Fabius Bile(aka “The Clonelord”), Dr. Manhattan, etc.

I am happy to agree that if we toss away a fundamental principle of science used in everything from understanding human nutrition and metabolism, through meteorology and plate tectonics, to atomic and astrophysics, you get to make up anything. Now you can finally say it was your prayers to the weather gods that made it rain, and that earthquakes and tsunamies come from society allowing homosexuals to marry and bare-breasted women on the beaches and streets. Just can’t wait for the weatherforecast on the creationism channel.

[Argon]

@John_Harshman

The pattern is what we expect from common descent

As with Ironic Design Theory as well, with one of its core claims being that nature was designed to make it impossible to demonstrate the existence or non-existence of the designer. Generic creationism doesn’t even come close to the powers of specification that Ironic Design possesses.

[colewd (Bill Cole)]

@John_Harshman

First, this is not a labeling fallacy. Nested hierarchy has a clear definition, and the data clearly fit that definition. Notice how I can account for almost every feature of the distribution by a single change on the standard tree. That’s not chance and it’s not just labeling. There is a real pattern here, and that pattern has a name.

If indeed there is a precise unambiguous definition and the pattern is an exact fit of that definition then you are not using a labeling fallacy. The test of this is you can remove the label and the argument will not change. You can do the same exercise with the label " Texas sharp shooter".

A single change or thousands of changes. Are you claiming the changes occurred at the same branch on a more detailed tree?

@John_Harshman

Second, of course special creation could create the pattern; special creation could create any pattern at all.

Any pattern does not create a functional animal. The set of genes must be able to create a functional animal from a zygote.

@John_Harshman

. But if we start accepting untestable hypotheses, we have abandoned any sort of science entirely. Is that really what you want?

@Winston_Ewert proposed a test of the pattern. On page 14 he shows a dependency graph of zebra fish mice and chickens.

@John_Harshman

No, it’s not possible. The pattern is what we expect from common descent. Interpreting it as creation would seem to require intentional deception on the part of the creator.

Yet the data can be arranged in a dependency graph with a better overall fit which hints at the design strategy of our Creator. Do you have a problem with @Winston_Ewert method of using a Bayesian model? If so why?