Behe's response to his Lehigh Colleagues

https://evolutionnews.org/2019/03/a-response-to-my-lehigh-colleagues-part-1/

“But, as the reviewers themselves insist, those results are all based on an unnatural situation — on the prevalence of degradative mutations in artificial environments — so why should we trust that the results reflect what would happen in nature? How can the reviewers with any consistency accept some of the lab results but not others?
It astounds me to see how quickly lab evolution researchers disavow the importance of their own life’s work when some outsider draws an unwelcome inference.”

This part of Behe’s response is breathtaking. Are we really expected to believe that Behe can’t fathom that laboratory experimentation can provide insights into natural evolution with caveats? That researchers can extol the importance of some parts of their research while accepting that other parts are difficult to extrapolate beyond the lab? This is very basic, undergrad-level stuff, and I really can’t bring myself to believe that Behe doesn’t understand it. The only possible conclusion is that Behe is deliberately muddying the waters for his lay-readers.

12 Likes

It is another example of black-white, all-or-nothing dichotomous thinking that underlies so much of creationism. Somehow this thing with nuances and limited areas of relevance and applicability just doesn’t enter their mind. Everything is seen in absolutes. An experiment in evolution is taken to constitute evolution in every possible aspect and environment, or not at all.

Somehow showing there are some caveats to an experiment is interpreted to mean researchers have fully and completely abandoned all aspects of their experiment under all possible circumstances.

It’s ridiculous. The man is completely blinded by this cognitive bias. And it’s very ironic, given this:
"It has been my experience that one very common way for opponents to try to discredit an argument is to exaggerate it, to ignore distinctions an author makes, and/or to change carefully qualified claims into bizarre absolutes.” - Michael Behe

3 Likes

It’s nice of Behe to distill the best examples from his book into a single paragraph though:

One big fly in their argument, however, is that they overlook the results from non-laboratory evolution that I give in the book. Every species that has been examined in sufficient detail so far shows the same pattern as seen in lab results. For example, I open the book with a discussion of polar bear evolution. About two-thirds to three-quarters of the most highly selected genes that separated the polar bear from the brown bear are estimated by computer methods to have experienced mutations that were functionally damaging. (Some other reviewers questioned this. I showed why they are mistaken here .) Similar results were seen for the woolly mammoth. Neither of those species evolved in the laboratory. Except for the sickle mutation (which itself is a desperate remedy), all mutations selected in the wild in humans for resistance to malaria are degradative. Dog breed evolution, which has been touted as a great stand-in for selection in the wild, is mostly degradative, and lots of breeds have health problems.

We have the polar bear example again, which I think at best could be described as “ambiguous”. Certainly what Behe says about 2/3rds to 3/4ths of mutations being predicted to be damaging based on the Polyphen predictions has been discredited.

Then we have the example of the woolly mammoth - Behe cites Lynch et al. (2015) for this. In his book, Behe says of this paper:

“Analysis showed, however, that of the approximately two thousand amino-acid residues found to be mutated in mammoths, about five hundred were likely to be damaging.64 Another three hundred changes couldn’t be decided, but a chunk of them too may be damaging. What’s more, a further twenty-six genes were shown to be seriously degraded, many of which (as with the polar bear) were involved in fat metabolism, critical in the extremely cold environments that the mammoth roamed.”

It seems that these figures of “five hundred” and “three hundred” changes being likely damaging/unknown but maybe damaging also come from Polyphen predictions, which are flawed for the same reason as Behe’s claims about the polar bear mutations are. I’m quite surprised how heavily Behe leans on Polyphen predictions to make his case (they form the basis of his top two go-to examples of real-world degradation-driven evolution in nature). It’s also quite telling that AFAIK, Behe hasn’t responded on the subject of the significance of Polyphen anywhere yet, despite writing responses to arguments that included criticisms of his interpretation of these predictions.

I think it’s fair to say that his last example of dog breeds is pretty inane, so that just leaves human mutations that provide resistance to malaria. Worth a new thread to discuss?

3 Likes

Good idea!!

In his book, Behe also relies of SIFT predictions of amino acid changes being damaging, despite the fact that in the Nature Protocols article describing the algorithm (Kumar et al. 2009), the authors say:

“‘DAMAGING’ means that the substitution is predicted to affect protein function”

Behe’s interpretation of these SIFT results are flawed in the exact same way as his interpretation of the Polyphen results.

4 Likes

In his book, Behe also relies of SIFT predictions of amino acid changes being damaging, despite the fact that in the Nature Protocols articles describing the algorithm (Kumar et al. 2009), the authors say:
“‘DAMAGING’ means that the substitution is predicted to affect protein function”
Behe’s interpretation of these SIFT results are flawed in the exact same way as his interpretation of the Polyphen results.

As someone who primarily does bioinformatics, this is really aggravating to me. These results simply are not a test of his hypothesis in the way that he’s claiming they are. Has Behe acknowledged this problem anywhere?

5 Likes

I agree with this comment as long as you can establish what the “way he is claiming it to be” is nailed down. This is a new hypothesis and the question is will it stand the test of time.

The genome is a sequence that builds complex animals. Observation shows many protein sequences appear to be optimized “wild type” sequences. All logic says that most random changes to optimized sequences will be neutral or degrading. I would expect that what he is observing will ultimately be true.

I realize this is very counter intuitive from the evolutionary perspective where you think random change and selection built them in the first place.

1 Like

Bill, as far as I can tell, your comment has nothing to do with what I actually wrote.

Do you agree that a computational analysis whose results can only be “no functional effect” or “deleterious” can not tell us much about how often adaptive changes (which by their nature must have some sort of functional effect) result in degrade function vs improved function?

1 Like

I honestly don’t know but would like to see this argument developed.

If you search for “polyphen” on ENV, no results turn up, so it seems he hasn’t commented on this problem in the recent burst of dialogue since his book was published.

Yes, but that doesn’t mean that most fixed changes (that correlate with a change in selection pressures) to previously optimised sequences (optimised to previous selection pressures) should be neutral or degrading.

2 Likes

I don’t see the logic here if you are starting from an optimized wild type sequence which is part of an optimized function.

At this point, he may have no choice. He has to double down on this obvious error. Look what he writes here:

. For example, I open the book with a discussion of polar bear evolution. About two-thirds to three-quarters of the most highly selected genes that separated the polar bear from the brown bear are estimated by computer methods to have experienced mutations that were functionally damaging. (Some other reviewers questioned this. I showed why they are mistaken here.)

The good thing about this move is it makes it very easy to explain to thoughtful students confused by this that Behe’s argument is false. He is either confused by, or is quote mining, these software algorithms. By doing this repeatedly, this seems to demonstrate this is central to his argument. This should make it clear to observers trying to make sense of this why biologists are not convinced.

Yes, he doubled down.

I wonder what he will do if the authors of these programs just change the output to “change of function”, as is more accurate?

3 Likes

The function is optimised according to a particular set of selection pressures. If those change, the previously optimal sequence may now be sub-optimal, so subsequent mutations can improve the function towards a new optimum.

1 Like

I honestly don’t know but would like to see this argument developed.

It’s pretty simple. Behe’s hypothesis seems to be that most adaptive changes result in degraded biochemical function. To support this thesis he relies on analyses that look for mutations with functional effects, all of which are considered to be harmful. There is no category of mutations that these analyses can classify as improved function (to do so would require much more work). So it’s not possible to determine the prevalence of harmful vs helpful mutations from these results.

2 Likes

Many of those classified as “damaging” might in fact not be damaging.

3 Likes

True. Which is yet another problem.

What observations are those? Don’t tell me you’re drawing a trend for the >90 million known protein sequences from a single data point.

But are you starting with an “optimized” wild type sequence? If it was “optimised”, why is it possible for mutations in it to be beneficial? In what sense is it then “optimized”?

2 Likes