Creationist genetic clocks

[Aggie]

This paper was published in Answers Research Journal around a month ago, and as far as I can tell there hasn't been any mainstream response to it yet. According to the paper (by Nathaniel T. Jeason), when one compares the diversity of mitochondrial DNA in non-African people to the mutation rates of mitochondrial DNA, it is most consistent with non-African people diverging from an ancestor around six thousand years ago.

The most recent issue of the Journal of Creation has a paper by Jeffrey Tomkins using different methods to reach a similar conclusion. This paper was featured on the journal's cover, so the journal editors evidently think it's a particularly big deal, but I haven't been able to find the paper online anywhere. If anyone here wants to look for it, the citation is: Tomkins, J. P. 2015. “Empirical Genetic Clocks Give Biblical Timelines.” Journal of Creation 29 (2): 3–5.

Now that most parts of Talk.Origins are no longer updated, we probably can't expect these papers to receive any response there. Can anyone at this forum find any flaws in the two papers' methods?

 

[Willtor]

Comparing mitochondrial DNA seems like comparing Y-chromosomes to me. Maybe I'm missing something? Mitochondrial Eve should be fairly recent, if peoples from across the globe have been in reasonably regular contact. What is it we're trying to measure, here?

 

[Aggie]

Comparing mitochondrial DNA seems like comparing Y-chromosomes to me. Maybe I'm missing something? Mitochondrial Eve should be fairly recent, if peoples from across the globe have been in reasonably regular contact. What is it we're trying to measure, here?

Measuring mutation rates in mitochondrial DNA is a fairly standard method for molecular clocks. (Some molecular clocks are calibrated using radiometric dates from fossils, but calibrating them based on present-day mutation rates is generally a more accurate method.) this (mainstream) paper gives an age of 165,000 to 188,000 years ago for Mitochondrial Eve.

This map shows the commonly-accepted divergence dates of human groups as measured using mitochondrial DNA. Mitochondrial DNA shows that indigenous Australians diverged from other groups 60-70 thousand years ago, and for Europeans it's 40-50-thousand years ago. Even Jeason's results are only for non-Africans, there's no way to reconcile them with these accepted dates from mitochondrial DNA if there isn't a problem with his methods.

 

[sfs]

Measuring mutation rates in mitochondrial DNA is a fairly standard method for molecular clocks. (Some molecular clocks are calibrated using radiometric dates from fossils, but calibrating them based on present-day mutation rates is generally a more accurate method.) this (mainstream) paper gives an age of 165,000 to 188,000 years ago for Mitochondrial Eve.

Calibrating with observed mutation rates isn't necessarily more accurate. Directly observed mutations include mildly deleterious mutations that will eventually be weeded out by purifying selection. This is a well-known phenomenon and easily observed; for example, we saw it in the virus in the recent Ebola outbreak, which had a markedly higher substitution rate over the first few weeks than over the first six months or so. Correcting for this requires a really good understanding of the substitution rate over different time ranges. The paper you cite above makes that kind of correction, but is using rates calibrated from the human-chimpanzee split (taken from this paper) -- unless I'm reading it wrong. The size of the effect at the single-generation level (needed for correcting mutation rates from single-generation pedigrees) is not easy to estimate, and could be quite large. The creationist paper deals with the issue by ignoring it.

 

[Willtor]

Measuring mutation rates in mitochondrial DNA is a fairly standard method for molecular clocks. (Some molecular clocks are calibrated using radiometric dates from fossils, but calibrating them based on present-day mutation rates is generally a more accurate method.) this (mainstream) paper gives an age of 165,000 to 188,000 years ago for Mitochondrial Eve.

This map shows the commonly-accepted divergence dates of human groups as measured using mitochondrial DNA. Mitochondrial DNA shows that indigenous Australians diverged from other groups 60-70 thousand years ago, and for Europeans it's 40-50-thousand years ago. Even Jeason's results are only for non-Africans, there's no way to reconcile them with these accepted dates from mitochondrial DNA if there isn't a problem with his methods.

I see. Mitochondrial Eve is much older than I realized.

 

[sfs]

I see. Mitochondrial Eve is much older than I realized.

The expectation for the age of mtEve is one-quarter the age of the nuclear genome, which would put it around a couple of hundred thousand years. The variance on that expectation is very large, however.

 

[sfs]

I'd also be a little concerned about another aspect of the creationist paper. Their mutation rate is based (in the end) entirely on homoplasic mutations (ones present in the entire set of mitochondria in an individual) in a single study, one that was not focusing on homoplasic mutations. The mutation rate comes from the 0.5% of homoplasic variants that differ between mother and daughter. The authors of that study estimate their sequencing error rate (false discovery rate) for heteroplasmies to be 2%, but don't provide an estimate for homoplasies. You have to be very confident in your sequencing to be sure that the error rate is much less than 0.5%. That's not an issue for the original paper, but if you're trying to use their results for estimating the mutation rate, you'd better understand the error rate.

 

[Aggie]

I have a few more questions about this paper, but I'm going to save them for now, because sfs has offered (via e-mail) to look for the Tomkins paper I mentioned in my OP. What I'd especially like to know is whether there's a common flaw that's causing all of these creationist genetic clocks to produce inaccurately short timelines, along the lines of how all of Tomkins' BLAST comparisons of the human and chimpanzee genomes produced excessively low percentages due to being ungapped.