tisdag 17 februari 2015

Richard III the Usurper King and Mitochondrial Eve

Here is an interesting tidbit for the creation-evolution debate, which concerns the creationist vs. the mainstream scientific takes on the true age of mitochondrial Eve. I may be the first to spot this particular connection, so I'm throwing it out there as a rather hasty blog post.

The infamous king Richard III of England.
Picture from Wikipedia.
I have previously written a blog post (in Swedish) about evidence and how it can be used to confirm or falsify hypotheses. In that blog post I mentioned the case of the missing and found remains of Richard III of England, the ill-reputed usurper and probable regicide (see Rex Factor Podcast). The historical chronicles and the modern archaeological case regarding Richard III are also abolutely fascinating stories all by themselves. In short: the recently rediscovered remains of Richard III were identified by comparisons with mitochondrial DNA from his skeleton and that of two modern-day matrilineal descendants from his sister. I originally used this only as an illustrative example of evidentiary support for a hypothesis, but during a discussion in the Evolution Fairytale Forum it occurred to me that mitochondrial data of this kind also has general relevance for estimations of mutation rates in mitochondrial DNA. Which brings us to the case of mitochondrial Eve. I am indebted to forum member Mike the Wiz for throwing mitochondrial Eve into the discussion (and probably not realizing in advance that by doing this he was biting off a lot more than he could chew).

Estimates of mutation rates (=base substitution rates) in mitochondrial DNA over time have been performed from various sources, based on combined palaeontological, achaeologial, historical and genetic evidence (see e.g. Rieux et al. Molecular Biology and Evolution 2014). Based on these rates, and the overall variation between different human mitochondrial DNA lineages from within and outside Africa (see Wikipedia, Oven and Kayser 2008, Endicott et al. 2009, and Mellars 2006), a time point for their divergence from the last matrilineal common ancestor of all humans can be calculated. These calculations by mainstream science seem to clock in at approximately 200 000 years ago, with the likely point of divergence somewhere in Africa.

However: other studies of mutation rates have yielded substantially higher estimates. These rates are obtained from modern comparisons of changes over a few generations, as determined from pedigrees or cell lines (see e.g. Parsons et al. Nature Genetics 1997; Howell et al. American Journal of Human Genetics 2003; Madrigal et al. American Journal of Physical Anthropology 2012). Creationists have leapt on this to claim that, based on these rates, a better time estimate for mitochondrial Eve would be around 6000-6500 years, bringing her close to their biblical estimate for the time of creation and the biblical Eve. The creationist take on the Mitochondrial Eve story is presented as a factual case (number 4) in Don Batten's list of 101 (supposed) evidence for a young age of the earth and the universe.

Note that the discrepancies between estimates over short-term and long-term time scales are considered genuine, and acknowledged by both mainstream and creationist publications. See an example here from the creationist publication Creaton ex Nihilo Technical Journal / Journal of Creation:

What are we to make of this? How should we then resolve these discrepancies between mutation rates over different time scales? Should either the short-term or the long-term rates be considered artefacts, or is there some explanatory model that would allow us to reconcile these different rates?

The most parsimonious explanation for the observed data, however, is probably that both rates are valid, but that differences between short-term and long-term mutation rates represent two different processes, acting over different time scales. The short-term rates represent the instantaneous frequency of mutations in the mitochondrial DNA molecule, whereas the long-term rates represent what remains in the gene pool after some filtering process has acted on the genetic variation over several generations and removed 90-95% of all mitochondrial DNA variants. Evidence that mitochondrial DNA is subjected to such purifying selection can be found in the differing substitution rates among bases in the first and second codon positions, and the third (synonymous) coding position, respectively (Rieux et al. 2014). This explanation allows us to harmonize the differing mutation rates, and provides strong support that the long-term rates would indeed be most relevant for comparisons over longer time scales. 

When confronted with this very reasonable solution, however, creationists would predictably respond with the usual canard about “observational” science. From the creationist point of view, the case appears shut: The short-term, higher rates should be preferred over long-term rates, ostensibly because they constitute "real" observations, rather than being calibrated estimates based on "evolutionary assumptions". In reality, both time frames are well supported by independent lines of evidence, and both estimates should therefore be considered factual. Anyone arguing this sensible point of view would probably have a hard time to break through the creationist trenches, fortified with rhetoric about evolutionary assumptions being a necessary presupposition for the long-term rates. 

Here is where good old (or bad old) Richard Plantagenet comes to our rescue, literally from the grave. The time frames for the Plantagenet dynasty and Richard’s short reign are not subject to any evolutionary assumptions, but constitute reliable historical facts that could hardly be questioned even by the most ardent creationist. As we have access to mitochondrial DNA from Richard himself, as well as from two independent matrilineal relatives, we can directly calculate a factual mutation rate over medium-range time scales, which can be compared with both short-term and long-term rates. 

The main creationist case for a mitochondrial Eve living roughly 6000 years ago comes from the study by Parsons et al. 1997, which found an unprecedented high rate of 1/33 generations per mtDNA molecule or 2.5 mutations per base pair per million years (2.5 x 10-6  per base per year). The forensic material from Richard and his modern relatives allows us to directly test the validity of the creationist approach to simply extrapolate short-term mutation rates over longer time frames. Have a look at the genealogy of Richard’s matrilineal relatives until the present day, from King et al. 2014:

There is a time span of roughly 550 years between the birth of Richard’s mother Cecily Neville and the modern descendants whose mitochondrial DNA was sequenced along with Richard’s. From this follows that the mitochondrial DNA molecule of each of the modern relatives will have accumulated mutations for roughly 550 years, compared with Richard’s template DNA. This means that we could expect the following number of mutations to have occurred:
16570 (the length of the human mtDNA) x 550 x 2.5 / 1000 000 = 22-23 mutations. 

In the King et al. 2014 study, the scientists sequenced the full mtDNA molecules of both Richard III and his modern relatives and found…wait for it…that the mitochondrial DNA of one relative was identical with Richard’s, whereas that of the second relative differed only by a single mutation! 

This is far from the expected value if the short-term rates were valid, even over a 550-year time span, and constitutes a slam dunk against the validity of the creationist approach to estimate longer time frames based on short-term mutation rates. 

The two mitochondrial DNA lineages of Richard’s modern relatives constituted a single lineage for roughly 115 years, until they split apart with the sisters Barbara and Everhilda Constable in the 1530:s. These two mitochondrial genomes have thus accumulated mutations for a total of 115 + 2 x 435 = 985 years in total. With a single mutation over this time, the mutational rate can be calculated as: 
1 / (1657 x 985) = 6 x 10-8 mutations per base per year. 

This is almost exactly 40 times lower than found by Parsons et al. 1997. It is also very close to the mutation rates of approximately 2-5 x 10-8 found by Rieux et al. 2014, based on different estimates using medium-to-long time frames. The data from the mitochondrial DNA of Richard III and his modern relatives thus constitutes a factual demonstration that mutation rates can approach the long-term rates already after only ca 500 years. This also provides a strong case that purifying selection or some other process is at work to filter out most mutations in the mitochondrial DNA after a very limited time span.
Unlike the long-range mutational rate estimates, the creationists cannot bring out their old canards against the case of Richard III. The estimated rates in this case rests on only three parameters, neither of which can be in reasonable doubt: 

1) The confirmed relationships between the skeleton identified as Richard III and the two modern matrilineal relatives. King et al. 2014 build a very strong cumulative case for this in their paper, based on genealogy, genetics, and historical descriptions of Richard’s battle wounds and the damage and deformities of the skeleton.
2) The genealogy and historical dates of the members of the Plantagenet dynasty and their descendants.
3) The sequences of the mitochondrial DNA molecules from Richard’s skeleton and those of two of his modern-day maternal relatives with strictly matrilineal descent, all of which have been confirmed by means of multiple sequencing. 

Thus, the genetic case for a mitochondrial Eve living more than 100 000 years ago stands stronger than ever. On the other hand: based on the approximately 1 mutations per 1000 years per mtDNA molecule found between Richard and his relatives, the creationist case for a most recent common mitochondrial ancestor only 6000 years ago is more absurd than ever. More cases like that of Richard III could probably be found, in order to strengthen the validity of this approach even further.

When we take a step back and survey the overall situation with regards to biogeographical age estimates for the human species and its populations, we find an amazing consistency between different types of evidence from historical sources, archaeology, paleontology, radiometric dating, and population genetics (Endicott et al. 2009; Rieux et al. 2014). The scientific case for age estimates of the human species does not rest on any “evolutionary presuppositions”, but constitutes a bottom-up reconstruction based on real historical data with a high degree of concordance between multiple independent lines of evidence. 

Let us contrast this scientific success story with the creationist approach to epistemological and scientific consistency. If you recall the original publications by Carl Wieland and CMI, the ostensible justification for preferring modern, short-term estimates for mitochondrial mutation rates was that these represent present-day observations, rather than shady and uncertain values calibrated by evolutionary assumptions. This may seem like an epistemically sound approach, but is it consistent with how the same scientists accept other modern, observed rates in other contexts? (see Age of Rocks for examples):

What is their take on modern, observed rates of radiometric decay?
And modern, observed amounts of C14 in the atmosphere (calibrated tens of thousands of years back by tree rings, sediment varves, and speleotherms)?
Or modern, observed rates of continental spread from the mid-Atlantic ridge?

In those cases they seem perfectly willing to accept any fairy-tale alternative for historical differences in rates, with little or no evidentiary support.

A more cynical, but entirely justified, interpretation of the creationist modus operandi is that they will always favour the alternative that supports their presuppositions and prior commitments to a literalist biblical interpretation. Verily, I say unto you, the level of hypocrisy routinely displayed by the "scientific" representatives of the major creationist organizations would make 
even a hardened pharisee blush.