Bartlett: Measuring Active Information in Biological Systems

Conversation Side Conversation
,
1

There’s a new article at Bio-Complexity.

I’m not sure how it ranks on the usual scale.

The abstract begins:

In computer search optimization theory, active information is a measurement of a search algorithm’s internal information as it relates to its problem space

I don’t think this is true. I can’t find any definition or references or publications about “active information” other than from the Discovery Institute. Nor dos the article have any references to articles about “active information”. There isn’t even a Wikipedia page on it. “Active information” appears to be just another ID buzzword of little relevance. At least this one has a formal definition.

The full text of the article begins:

Biological evolution operates in at least two well-known modes—either semi-directionalized, where the outputs of evolution are correlated with the selection pressures the organisms face [1], or as a non-directionalized drift, as neutral theory describes [2]. Historically, research into evolution has focused on the ability of natural selection to keep beneficial mutants in the population, and not how they originate to begin with.

It’s hard to know whether the first sentence qualifies as an error, fallacy or falsehood, because the author doesn’t seem to know the meaning of the words he’s using. Evolution may be semi-directional or non-directional, but it isn’t directionalized (unless Bartlett thinks its guided by the hand of God). It’s even arguable whether evolution is (semi-)directional. Bartlett’s first reference is to the 1859 Origin, which doesn’t mention direction (other than by artificial selection), selection pressure, correlated outputs or even “evolution”. As for “the outputs of evolution are correlated with the selection pressures the organisms face”, Darwin gave examples of linked traits where the “outputs of evolution” are not correlated with any selection pressure. Darwin also devoted space in Origin to the origin of variation, albeit to admit ignorance, and mutations have been studied since before the discovery of DNA, mutations to which have been studied for more than 50 years, so Bartlett’s claim about research into evolution historically may or may not be false depending on how “focussed” is interpreted.

Bartlett continues:

Recent work in evolutionary theory, especially in evolutionary developmental biology, has led to the realization that the inputs to evolution (i.e. the evolutionary paths that organisms are endogenously predisposed to take and the existing developmental pathways that canalize these changes into useful phenotypes) are just as important as the processes of mutation and selection themselves.

This is more bafflegab. Organisms are not endogenously predisposed to take particular evolutionary paths. Evolutionary paths are not the inputs to evolution. Developmental pathways don’t “canalize” developmental changes (that seems to be Bartlett trying but failing to impress with vocabulary he’s read but not understood*) and don’t really channel them either. Bartlett appears to be another IDer who, like Behe, mixes up development with evolution of development and evolution of form.

The rest of the article isn’t much better. It’s based on the usual misrepresentation of evolution as a search for specific mutations, and consequently fails completely to connect “active information” to anything biological.

So, how does this article scale? The abstract seems to score a zero, but that’s based on not finding any non-DI use of “active information”. The body might score anywhere from zero to two, depending on how Bartlett’s Humpty-Dumptyism is interpreted.

Thoughts?

*“Canalisation” has been used to describe how physical contraints of both organisms and the environment lead different species along the same broad evolutionary pathways, e.g. streamlining of ocean predators, or island dwarfism. It has nothing to do with developmental pathways.

1 Like
2

That isn’t true. In the literature, especially the paleontological literature, it’s been used to refer to the stability of development in the face of environmental influences and also of mutations. If you google “developmental canalization” you will find plenty of references. It’s been used to explain the (supposed) reduction of morphological disparity after the Cambrian explosion, for example. But Bartlett, in the quotes you provide, is still spouting gibberish.

2 Likes
3

For example:

Canalization, Developmental Stability, and Morphological Integration in Primate Limbs

4

Further extracts from the article:

Measuring active information is measuring the information that the genome (as it presently stands) has about likely beneficial future configurations.

That’s zero, surely?

What active information measures is the alignment of the genome itself to the problem of finding viable genetic solutions to selection pressures.

Hmmm…

While measuring active information is fairly straightforward in digital systems…

But not straightforward enough for Bartlett to actually have done it, which renders his claim very empty indeed.

What we would like is to let organisms produce their own mutations, measure the success rate of the organism’s experiments, and then let the experimenter produce random mutations, and measure the success rate of those experiments.

But would the experimenter use the same rates and probabilities of different types of mutation occurring as happens biologically? If so, then the “success rate” should be the same. If not, the results won’t mean anything.

Active information is, in actuality, a relative measurement, similar to decibels.

YADA (Yet Another Definition of “Active information”)

In general, the purpose of calculating active information is to see if the cell has mutational mechanisms geared to solving the given evolutionary problem presented to it

Effective mutation rates are dependent on an organism’s ability to detect and repair genome copying errors, which may itself be environmentally influenced. So yes, cells do have such mechanisms, and those mechanisms have evolved and continue to evolve.

Active information can also be important for bioengineering. Knowing the types of problems different organisms are geared to solve will help in determining the likely future evolutionary paths of organisms.

As well as the conflation here between organisms solving problems vs organisms evolving, that can already be done without “active information”. Active information is just another probabilistic metric that is irrelevant to biological evolution.

1 Like
5

Hey, I went to grad school with him. UC rules!

1 Like
6

Thanks John - that’s a different use of “canalisation” than the one of which I was aware. I learn something every day here.

1 Like
7

Others before us have dealt with this, let’s ask them …

Paging @Joe_Felsenstein and … Tom English … ??? … I though Tom was here already.

ETA: Tom English: Review of Evolutionary Informatics (Marks et al)
ETAA: Why ID Advocates Downplay Our Disagreement With Them

ETAM: An index of all of Joe’s posts on Panda’s Thunb! :slight_smile:Felsenstein posts at Panda's Thumb

8

Which, coincidentally or not, happens to also be the amount of information that Bio-Complexity (as it presently stands) has about likely beneficial additions to our knowledge of biology.

3 Likes
9

This is very poor communication I have to say. In trying to make sense of that sentence I get the impression that the author is trying to say that there is some degree to which the genome of an organism is likely to change in a beneficial way in some environment, and that this degree to which it is likely to do so, can be quantified and measured. And that the name of the unit of quantity is “active information”.

But isn’t that essentially just evolvability? Or maybe plasticity? So he’s sort of coming up with some sort of unit of measurement of evolvability?

2 Likes
10

OK, looking at equation (3), this is nearly the equation for a Likelihood Ratio statistic, differing only by a constant (~1.4). Bartlett goes on to do interpretation and whatnot, but he is still committing the equivalent of interpreting the magnitude of a p-value. Under the same assumption of randomness, equation (3) is a random variable is the same way a p-value is a random variable. No, a p-value does not have a “strength of difference” interpretation.

This might have a theoretical “non-centrality” interpretation, related to the statistical power to reject the null hypothesis of “non-centrality equals zero”. Measuring this quantity from data it only a way to restate the p-value of the test.

Other than that, he entirely ignores selection and the efficiency of genetic search, which is darned good, O[n*log(n)].

2 Likes
11

23 posts were merged into an existing topic: Side comments on Bartlett: Measuring Active Information

13

I’ve got the paper; will hold off until I get a chance to read it – quite busy today and tomorrow.

2 Likes
16

I don’t say any of them are wrong. It’s the swaggering “I know how evolution works, and I know this guy is wrong, as all ID proponents are wrong because they don’t know what they’re talking about” attitude that I object to – an attitude all too common on this site, which is why I hardly post here any more, or even read what’s posted here.

I haven’t analyzed Bartlett’s article yet, but I know him personally, and I know his qualifications, and he has advanced training in areas relevant to his article; whereas, based on Roy’s years of commenting here, I can’t see that he has training in anything at all, except studying opinions posted on blog sites that discuss evolution. I’m just asking for the grounds of Roy’s confidence that he knows what he’s talking about and Bartlett doesn’t. If he doesn’t want to answer, he doesn’t have to. Just registering that I don’t accept Roy as an authority on evolution, any more than I accept him as an authority on the history of ideas.

But then, very few people who post here are specialists in evolutionary theory (as opposed to biologists, biochemists, or other people with a keen layman’s interest in evolution) – I think you and Joe Felsenstein are the only two people posting here with Ph.D.s in the life sciences specializing in evolutionary theory. But you aren’t the only two who represent themselves as experts in it. In fact, Joe and you are less aggressive-sounding in the expression of your opinions than are a number of champions of evolution here with little to no formal training in the area. Why there has to be this swagger and bravado all the time is beyond me. Does it come automatically with the science training, or are we getting a non-representative sampling of scientists here, i.e., is it the case that only the angriest and most biased people with science training tend to argue on blog sites about evolution, and that the moderate and fair-minded ones aren’t usually found in such places?

21

Yes, it’s possible. But where that’s the case, the person usually sounds just as competent, both on matters of detail and on general exposition of the field, as the trained person. I have not found that with Roy’s posts, which strike me as pastiches of borrowed thought and arguments as opposed to an internalized mastery of first principles.

I’ve seen no evidence that Roy reads any of the literature; he repeats talking points.

I suppose that naked contradiction followed by “Bartlett doesn’t understand what he’s talking about” (the gist of Roy’s criticism) “makes sense”, but it isn’t much of an argument. It’s just nay-saying.

It my understanding that he has training in engineering, mathematics and computer science relevant to the various models he talks about and arguments he makes. He doesn’t have to be a master of every aspect of evolutionary theory to understand one small aspect of it well. The question is whether Roy has mastery of even one small aspect, or merely repeats the popularized defenses of the status quo against ID. I suspect it’s the latter, as I’ve seen nothing like personal mastery of any area of any science shown in any of his posts (as opposed to, say, your knowledge of animal classification, which appears to show intimate knowledge of the field).

I’ve answered you out of politeness, because you addressed me, but I have to get back to work now, so I’m exiting this thread.

28

Yep. The Blyth Institute is located in Johnny B’s home in Oklahoma, and he seems to be the only member. You’d think he could at least get a green screen shot of a science lab to match BIO-Complexity’s state-of-the-art ID research facility. :slightly_smiling_face:

4 Likes
34

From Jonathan Bartlett’s, Technical Lead at ITX Corp, linkedin page listing for education, he has Union High School following with four years at Oklahoma Baptist University, with the (presumed) degree unspecified. From other articles, he also attended Phillips Theological Seminary, and has moved from an old earth to a young earth creationist camp.

Jonathan Bartlett … said he once made fun of people who believe in a young Earth, but now he is one.

Bartlett was motivated to study biology and other disciplines after his first son was born with a rare genetic disorder and died a few years later. That study changed his mind about the age of the Earth.

As a graduate student at Oklahoma Baptist University and later a theology student at Phillips Theological Seminary in Tulsa, belief in an old Earth was his default position, Bartlett said.

But the more he studied it, he said, the more he came to believe that arguments for an ancient Earth were often circular; people in that camp had their model of what is true and tried to make the facts fit into it. Some fit nicely, others did not, he said.

“When I look at a larger-scale picture of geology, it supports young Earth creation,” he said.

Debate on age of the earth heating up

2 Likes
39

I have created a side comments thread and moved a number of comments. It was not possible to completely separate comments I thought to be relevant from the squabbling, so I had to make a judgment call:

  1. Eddie’s personal experience with Bartlett is relevant.
  2. The origin/owner of the Blythe Institute is relevant.

The remaining personal attacks or snark should be disregarded, and I will police the thread to keep that from breaking out again. /fnord

ETA: There were a few comments moved that might be acceptable criticism, but such judgement calls are difficult. If you replied to an inappropriate comment it gets moved along with the rest - that’s all I have time to do.

1 Like
40

The first thing that comes to my mind is “nature has already done these experiments”. The genomes of modern species are a record of accumulated random mutations, and I think they would be well served using that data. There’s tons of information on the human genome and human biology, and there are several other ape genomes to compare to.

Another problem I see is the different types of mutations. From the paper:

My gut reaction is that it may not be proper to compare the substitution rate with the indel rate. In the Lenski experiment, it was a recombination event that moved a promoter in front of the citrase gene. So how does one model that, and how was the recombination rate increased or decreased in different hypermutator lineages? What are the chances that a functional promoter that is active in aerobic conditions would find its way in front of the citrase gene?

2 Likes
41

Let me restore some relevant information that’s been excised:

Bartlett is a YEC with no expertise in evolutionary biology. The Blyth Institute is his personal project, apparently run from his house. And his writing is not very coherent. That is, it’s hard to figure out, based on his words, what he means, or in fact whether he means anything.

5 Likes
42

Can anyone explain what this means? How is active information measured? Why is information being compared to, apparently, mutation rate?

1 Like