James Tour Update

I attended a talk by James Tour last night at a large, highly conservative church in my area. I’ll start with a couple of good points. I was glad to see him mention @swamidass and mention his work on GAE. He praised the book and highly recommended it to the audience. Additionally, to his credit, toward the end of his talk, he did mention an incident (without using names) when he accused a researcher of lying. He admitted fault and seemed sincere in his wish that he hadn’t used such strong language. He also mentioned a conversation he had with the other party over the phone that was quite gracious.


He proceeded to attack Szostak’s reasoning that was discussed in the phone call. I wish he would have left it alone, because that definitely added a rather bitter aftertaste to what sounded like heartfelt regret over the incident.

He also seemed to have his finger on the heartbeat of his audience and definitely geared his talk to what they wanted to hear. Several of his points could have been wrapped up as “Scientists make a ridiculous guess, publish some research that doesn’t show what they say it shows, whip up science media to exaggerate claims, take advantage of an ignorant public, and cram these results into textbooks.” This sequence of events was portrayed both for OoL research and for evolution, in general. I found this incredibly ironic as he was using tactics very similar to what he claimed the rest of the scientific community uses.

The meat of his talk was in stark contrast to that of a talk he gave about 13 months ago to scientists and physicians at the MD Anderson Cancer Center. After the MDACC talk, he was asked directly if he believed that the theory of evolution was the best explanation for what is not observable in biology today. He answered that evolution was “the only ballpark in town”, which I thought at the time was an odd way to put it, but it apparently concealed that he has serious doubts about evolutionary theory. Last night, he went so far as to mention Mary Schweitzer’s research and expressed doubt about the preservation of soft tissue. I don’t know if he intentionally sounded YEC-ish to a YEC audience, or if he really has YEC-ish leanings.

What was particularly upsetting was the rather cagey way in which he discussed human/chimp genome similarity and the ENCODE project. He claimed, in front of hundreds of people, that the 95% similarity referred only to the 1.5% of the genome that contains genes. He stated (possibly technically accurately, but incredibly misleading) that scientists were finding more and more dissimilarity in intergenic regions, but did not address the extremely high degree of similarity in those intergenic regions. Essentially, he pulled off the impressive trick of taking a jab at the term “junk DNA”, while simultaneously misleading the audience to believe that non-coding DNA was completely mismatched.

Outside of the GAE reference and the admission of fault for the Szostak attack, I could not have been more disappointed.


No, that isn’t technically accurate. 1.5% of the genome contains protein-coding exons, and they’re 99.5% similar between humans and chimps. 95% similarity refers to an average over all the genome if you count each base of each indel as a difference, which is a silly way to do it. Of course, very few of those indels are inside protein-coding exons. The more usual and more instructive figure is 98.7% similarity, which is the average base match of aligned sequences, not counting indels as aligned. It applies mostly to intergenic regions, because those are most of the genome. Sanford sounds both misinformed and (it’s hard to avoid this conclusion) dishonest.


Thanks for the clarification. His explanation certainly didn’t match up with the publications I’ve read.

The “possibly technically accurate” part I was referring to was that scientists were finding more dissimilarity in intergenic regions. Technically, as more sequencing information is confirmed for highly repetitive sequences or other regions that present difficulties, could it be construed as accurate (although misleading)?

Don’t get me wrong, Tour (not Sanford) definitely sounded misinformed. I’m just not quite ready to label that as dishonest.

1 Like

Sure. Short repeats, by their nature, will be harder to align than average sequences. There will be a lot of slippage and not much information to guide alignment even without it.


He seems to have been tailoring his message to his audience to such a degree that it’s hard to avoid that label. Imagine if the audiences had been switched.

1 Like

Yeah, I can’t imagine him making those types of claims in front of an audience of people that had knowledge of the subject. Indeed, he didn’t in the previous talk I attended a year ago. Like I said, I was rather disappointed.

98.8%, assuming you’re referring to the result from the initial chimpanzee genome paper. (The single-base difference rate was 1.23%. For some reason a lot of people seem to round that to 1.3%.)

1 Like

Let’s compromise on 98.77%. And yes, I was referring to the chimp paper.


I think a Vista track made using a genome browser would put this to rest, in a way a lay audience can understand. The differences from humans as one goes from chimp, to mice, to birds, to fish are so very obvious that it is hard to deny or avoid the proper conclusion.


Creationists love to argue that the figure for human/chimp similarity is Not That High. By quibbling about indels you can move it up or down. But what they never point out is that no matter what you do to raise or lower the figure, the figures for Human/Gorilla, Human/Orang, and Human/Gibbon track along with those for Human/Chimp, with Human/Chimp always highest, Human/Gorilla a bit lower, Human/Orang lower, and Human/Gibbon even lower.


Yes, and as @swamidass has pointed out in several places, as long as the methodology is consistent, the human/chimp genome comparison is always closer than the mouse/rat genome comparison.


The way the obfuscation trick works is by excluding controls. That makes the number meaningless to interpret.

In Dr Tours case he trusted the wrong biologists, and didn’t check for himself. This is unfortunate. His reputation in chemistry towers over the biologists he decided to trust in this case.

1 Like

I don’t know how he decided who he could trust, but that decision-making process really doesn’t make any sense to me. It isn’t like he doesn’t have qualified experts on his own campus or a computational biologist friend!

1 Like

I have told him my opinion of this. I am not sure he grasped my point. Ultimately, he is his own man, and has a right to say what he will, bearing the rewards and consequences alike of his public statements.

1 Like

Which biologists did he trust? What biologist of any sort would ever tell him that 95% similarity is measured over protein-coding exons?

1 Like

I’d say it crosses the line from ‘unfortunate’ to ‘unprofessional’.


I understand your comments, and would not dispute you, which biologists? I sure reasonable guesses could be made from public evidence but I request confirmatory questions are not directed at me.

Come now. You made a statement about who he trusted, and I’m only asking for you to clarify. If you didn’t want to say you shouldn’t have mentioned it.


he went so far as to mention Mary Schweitzer’s research and expressed doubt about the preservation of soft tissue.

That’s great news! Thanks!

One illustration - one random section of the human genome, and similarity tracks shown beneath.

The different genes/transcripts are shown at the top. The similarity tracks are, from top to bottom, chimp, gorilla, orangutan, gibbon, rhesus monkey, mouse, rat, dog, armadillo, opossum, platypus, chicken, frog, zebrafish, and lamprey. Darker shading means greater nucleotide identify.


Is this from the UCSB Genome Browser? I’ve tinkered a TINY bit with it, and the output looks similar.