Part 2 of Polar Bear Seminar


Part 2 of DI’s Polar Bear Seminar.

(S. Joshua Swamidass) #2

@art and @nlents. Are they just ignoring the fact that the authors concluded that the mutations were not damaging?


Here is part 3

(Nathan H. Lents) #4

I just read part 2, quickly but completely. There’s nothing new at all, they just restate the original argument, but sort of flesh it out a bit more? It doesn’t answer the critiques at all, it just repeats the original claims more slowly. So did they “engage” with the criticisms? Only if you count, “Nuh uh!” as engaging.

So there’s nothing new here to respond to. I’m not going to take their approach of just repeating the original claims more slowly. I think it was the right thing to do for scientists to rebut Behe’s claims in Darwin Devolves, and I now think it’s the right thing to do to move on and leave them to shout those claims over and over into the wind.

(Nathan H. Lents) #5

Also, after thinking about it, I am not joining @Art in ascribing authorship to Behe and everyone else for these unsigned posts. Unsigned posts are a way for Behe (or whomever) to attempt credit for answering critiques but avoid responsibility for any (additional) errors and poor reasoning. Maybe these posts are a team effort and that’s why they’re unsigned? Team efforts are great. We do them in science all the time. They’re called research articles. All the contributors are listed as authors and therefore take a measure of responsibility for the claims and content. These unsigned posts are cowardly. They also contain a lot of smears, but that’s the DI way and most posts, singed or not, have them. If it’s Trollhoffer writing, that’s all it will have.

(Arthur Hunt) #6

Yeah, pretty much. At least we know that calling Behe out on his misrepresentation of the supplemental table from Liu et al. made a real mark. It cannot be rationalized away.

(Nathan H. Lents) #7

yeah, I noticed that, too. I think including the full table is the closest they will get to admitting that it was deceptive to doctor the chart in the original hissy fit response in which he called us “incompetent.” We’re so incompetent that they’re issuing a five-part clarification in addition to the several other posts that discuss this? (Or did he just call me incompetent and you were “some other guy?” I can’t keep track.)

(Arthur Hunt) #8

That’s fine. We can agree to disagree on our approaches towards EN authorship and attribution.

1 Like
(Nathan H. Lents) #9

without calling each other incompetent???

(Arthur Hunt) #10

If the Red Sox don’t start winning, that’s about the nicest thing that will fly off my keyboard.

1 Like
(S. Joshua Swamidass) #11

I agree. Lenski was really clear too, showing how they just don’t understand polyphen, and quote mined the article. At least it will be easy for open minded students to see that they are wrong. That is the bright spot.

1 Like

What is it with all these references to football teams?


In my view, the problem with Behe’s argument is a change in language. PolyPhen-2 predicts that a mutation will be damagin. Behe claims that they are damaging. Predictions aren’t facts, they are just predictions. You have to follow up with additional work to determine if the predictions are correct.

On top of that, what does PolyPhen mean by “damaging”? It simply means a change in function. It doesn’t mean what Behe seems to think it means.

From the article:

In other words, Behe has leaned out over his skis. He has stretched this data beyond what it can currently show. If you are looking to overthrow one of the most well supported and widely accepted theories in science, you need something more than shaky functional predictions that are known to give false predictions.


For reference, you can find the ENSMBL page for human APOB here:;g=ENSG00000084674;r=2:21001429-21044073

From what I can see, the two most common missense variants in the human population are at the alanine at position 618 and the proline at position 2739. They have frequencies of .485 and .366 respectively. Both are given a probability of being damaging of 0.999 by PolyPhen. Given the frequencies of these alleles, and Behe’s comments about “damaging” alleles resulting in dysregulation of cholesterol, we would have to conclude that a massive portion of the human population should either have hypercholesterolemia or hypocholesterolemia. Is this the case? Nope.

Could it be that PolyPhen makes bad predictions? Yep.

1 Like

I will also plug my thread on APOB structure and function:

Of interest is the relative cholesterol levels in both polar bears and brown bears. From what I can see, both brown bears and polar bears have comparable cholesterol numbers. From the other thread:

If damaged APOB results in hyper or hypocholesterolemia per Behe’s claims, then why isn’t that seen in those figures? Behe or the DI needs to explain this.

(Brandon) #17

From polyphen-2’s webpage (Overview[PolyPhen-2 Wiki]):

The function of the polar bear APOB is not being directly predicted. The prediction has to be true for the human protein first. Then the effect in the human protein has to be accurately apply to the polar bear gene.

I’m not sure how PolyPhen works fully, and I haven’t tried to use it myself. However, it seems to me that since the polar bear genome is now deposited in NCBI, a repeat assessment of the polar bear mutations by PolyPhen today in 2019 would detect the polar bear APOB as a homolog through a BLAST search. The 2014 Cell paper was publish in May and the genome was release in Aug 2014 (PREDICTED: apolipoprotein B-100 [Ursus maritimus] - Protein - NCBI). So the PolyPhen analysis in their paper would not have found the polar bear as a homologous protein back then (because it wasn’t available). However, the polar bear APOB is in NCBI now so the very same polar bear substitutions which were predicted to be damaging in 2014 might no longer be predicted to be damaging because they would be found within “the spectrum of substitutions observed at the position in the family of homologous proteins” of currently available genomes.

Honestly, I think the authors of the cell paper were a little sloppy and did a little sleight of hand in connecting the effect of the mutation in the human gene to the polar bear. I don’t doubt their conclusions, but as a biochemist (I never trust biologists who do biochemistry), I have a criticism about their analysis here.

1 Like
(Curtis Henderson) #18

I give Klinghoffer at least a modicum of credit for actually using his name on his pieces.

1 Like
(S. Joshua Swamidass) #19

Klinghoffer is actually a pretty nice guy. I still remember the beer we shared in Seattle. I think it’s a game to him :slight_smile: . Klinghoffer, if you are reading along, come grab beer with me in STL. It is on me.

1 Like
(Nathan H. Lents) #20

But that’s why the whole thing is so perplexing. Do they want to be taken seriously or not?

1 Like
(Nathan H. Lents) #21

That may be true, but I think it’s pretty forgivable because this wasn’t really the main thrust of the paper and I don’t think they envisioned that this particular part would be under such scrutiny.

1 Like