Michael Egnor, Reverse Engineering and Genetic Algorithms

The story begins with Time magazine and Michael Lemonick, who wrote an entry for their Eye on Science blog which critizized the Discovery Institute’s silly propaganda piece, their list of “dissenters from Darwinism”. (The National Center for Science Education maintains a list of non-dissenters all of whom are named Steve which is longer than the DI’s list has ever managed to be.) PZ Myers explains what happened next, and how Michael Egnor advances onto the stage:

Now here’s the funny thing: the distinguished brain surgeon Dr Michael Egnor shows up in the comments and spouts the usual boilerplate claptrap we hear from these guys all the time: oh, he was a ‘Darwinist’ once upon a time, but then he was convinced by the complexity of the cell that ‘Darwinism’ had a problem. Sweet Jebus, but one thing that pisses me off is ninnies who equate complexity with design; random processes are excellent tools for making things extravagantly complex.

Michael Egnor has become the latest creationist darling, spewing the same sort of anti-scientific canards which have been refuted for years, if not decades. (Browse the TalkOrigins FAQ for a catalog of these memes. Usually, “frequently asked questions” are asked pretty frequently, but creationists do an astonishing job of asking them all the time and indeed almost never asking anything else!) Egnor‘s writing for the DI’s “Media Complaints Division” has irritated so many people in the reality-based community that the first two pages of Google hits on his name are, with two exceptions, scathing blog entries.

Anyway, Michael Egnor‘s latest ramble gave me an idea for a fun project to try; for details, read below the fold.

Not quite three weeks ago, a surgical oncologist challenged Michael Egnor to do the following:

1. Explain, specifically, how the design inference is “of great value” in medicine. Please support your description with concrete examples, preferably documented in the biomedical peer-reviewed literature that show how the design inference makes one a better physician or contributes to treatments for disease.

2. Explain, specifically, how the design inference has been of “enormous help in scientific research in general and medical research.” Please support your description with concrete examples documented in the biomedical peer-reviewed literature that show “best medical research” that is based on the “search for patterns recognizable as design.” In other words, show us examples of medical research either based on or strongly influenced by the design inference, and how the design inference led to or facilitated the discovery of a better treatment for a disease or a better understanding of the pathophysiology behind a disease.

Finally, after keeping us all on tenterhooks, Egnor responded!

The natural place to start showing examples of the inference to design in medical research is the seminal biological discovery of the 20th Century—Watson’s and Crick’s discovery of the structure of DNA.

As amply elaborated elsewhere — teh stupid, it burns!

Not only did Egnor base his rhetoric upon a single paragraph in a Wikipedia article (which has since been deleted as irrelevant to the subject), it’s just. . . well. . . incoherent on its own merits. He makes a big deal about “reverse engineering”, which in Creationist Bizarro World implies some pretty odd things. Orac summarizes:

It’s amazingly simple once you see the light, particularly if you have a beer or two (or three or four) first. What Dr. Egnor has done is to preemptively inoculate himself and his version of “intelligent design” against any criticism or any need to provide any real evidence for design and at the same time redefined science, all based on the semantics of a single sentence in a definition in Wikipedia!

The mind boggles.

Think about it. If Dr. Egnor is correct and all science that seeks to understand the structure and function of biological molecules is nothing more than “reverse engineering” of “biological machines” (a truly unfortunate choice of words by the Wikipedian who contributed that sentence to the article), then, in Dr. Egnor’s mind, that must mean only one thing: the structures being “reverse-engineered” must have been designed. After all you only “reverse engineer” machines, or computer code, or devices, and all of those things are designed.

That’s it. That’s the core of his argument. That’s really all there is to it.

Down in the comments, Torbjörn Larsson notes,

If Egnor wants to play word games, it is really word against word of course. With a little bit of effort, and not because it would be practical but possible, we could use evolutionary methods such as GA’s to solve reverse engineering problems, for example to decide values for parameters.

In fact, Egnor’s own reference mentions black-box reverse engineering, when the only thing known is the observed behavior but not the specific mechanism. (For example software without source code.) A design inference won’t help here, so Egnor can’t claim it is a complete description of all cases of reverse engineering.

And, you know, that got me thinking.

We could even in principle use genetic algorithms to figure out the structure of DNA itself! Let a “gene” in the computer’s memory be the spatial locations of molecular units: sugars, phosphates, purines, pyrimidines — the small molecules which Franklin, Pauling et al. knew were the constituents of DNA. Then iterate the GA, using as fitness function a comparison between a calculated X-ray diffraction pattern and the X-ray images taken experimentally.

Rosalind Franklin's X-ray picture of DNAThis is a bit of a perspective shift. Instead of thinking like sane people do about DNA carrying genes, we’re considering an abstract sort of gene which defines the shape of a hypothetical DNA structure. Imagine tossing out a thousand random guesses about what DNA looks like. For each guess, we could calculate what the X-ray diffraction pattern would look like given that particular molecular structure. Most of the time, it won’t look anything like the X-ray pictures we take in the laboratory, but a few of them will by happy accident look a little more like the real thing. This slight preference becomes the starting point for selection. We let our ideas breed, giving favor to those which perform best. The irresistible logic of Darwin goes to work.

If we’d invented fast and cheap computers before we knew about DNA — say, in some parallel Sliders world or steampunk fantasy where computers happened five decades sooner — this might well be how scientists would have tried to solve DNA. It requires much less cleverness, and correspondingly more computer time.

Has anyone tried a trick along these lines? Google Scholar returns oodles of hits “for genetic algorithm x-ray diffraction”, including a 2000 paper by Chacon et al. which reports finding structures at ~2 nm resolution for myoglobin, troponin C, spermadhesin PSP-I/PSP-II, chymotrypsinogen A, superoxide dismutase, ovalbumin, tubulin, nitrite reductase and catalase using a GA which explores a discrete search space of sphere-packing configurations. I did an undergrad thesis on (in part) extracting scattering cross-section information about gas molecules using genetic algorithms, actually; the general idea is much the same, except that we’d be considering X-ray photons hitting molecules instead of electrons doing so, and the molecules would be significantly larger (increasing the dimensionality of the search space). Still, if people have been able to get the technique to work with proteins, I see no theoretical reason why one couldn’t use a GA to recover, or rather re-discover, the DNA double helix.

Incidentally, there’s a good chance that Francis Crick was experimenting with LSD while he and Watson were trying to beat Linus Pauling to the DNA structure. Are we now forced to say that LSD is essential in understanding the “design inference”, that one cannot know the Great Designer without LSD, or perhaps that LSD is God Himself?

UPDATE: For those coming here from the Skeptic’s Circle, we have a follow-up post here.

2 thoughts on “Michael Egnor, Reverse Engineering and Genetic Algorithms”

  1. The trouble is that as far as the Egnors of this world are concerned, GAs — and indeed any artificial system that demonstrates the efficacy of variation+selection — is itself the product of design and therefore cannot possibly constitute evidence in favour of evolution. Whoever wrote the code *obviously* secretly included a complete design of the end product they were seeking. It stands to reason.

Comments are closed.