December 29, 2009

Imperfection of the Molecular Clock of Hominid Mitochondria (Loogväli et al. 2009)

Another paper which shows that post-Last Ice Age mtDNA evolution did not proceed "slowly", i.e., with a rate similar to that inferred from human-chimpanzee divergence. As the authors note, the mechanism for the "slow" mutation rate, i.e. purifying selection of deleterious mtDNA mutations was "sluggish".

Note also how post-Last Ice Age population expansions are responsible for the weakness of purifying selection: as populations expand, even slightly deleterious alleles (and even non-deleterious ones) have a higher chance of surviving in a population. Genetic diversity increases faster in an expanding than in a static population.

From the paper:
In conclusion, human mitochondrial DNA clock is time-dependent mainly because of the time-dependence of purifying selection. There is also evidence that purifying selection has been counteracted by other population genetic factors during the course of human history. Our results imply that the proportion of synonymous substitutions has alternated between growth and decrease. This interpretation is strengthened given the shape of the human mtDNA tree, which reflects the bottlenecks and subsequent population expansions associated with the out-of-Africa migration and the hectic climatic conditions of the last glacial period. The wavy growth of the proportion of synonymous substitutions implies biases in the published correction of the mtDNA clock, which assumed a monotonic growth curve [15]. Therefore, the clock of synonymous substitutions should be preferred. In addition, it seems that a good consensus has been achieved on the rate of accumulation of synonymous substitutions in human mtDNA, which applies at the population as well as the interspecies level (this study, [15]).
[15] is Soares et al. (see below)

Related:
PLoS ONE doi:10.1371/journal.pone.0008260

Explaining the Imperfection of the Molecular Clock of Hominid Mitochondria

Eva-Liis Loogväli et al.

Abstract

The molecular clock of mitochondrial DNA has been extensively used to date various genetic events. However, its substitution rate among humans appears to be higher than rates inferred from human-chimpanzee comparisons, limiting the potential of interspecies clock calibrations for intraspecific dating. It is not well understood how and why the substitution rate accelerates. We have analyzed a phylogenetic tree of 3057 publicly available human mitochondrial DNA coding region sequences for changes in the ratios of mutations belonging to different functional classes. The proportion of non-synonymous and RNA genes substitutions has reduced over hundreds of thousands of years. The highest mutation ratios corresponding to fast acceleration in the apparent substitution rate of the coding sequence have occurred after the end of the Last Ice Age. We recalibrate the molecular clock of human mtDNA as 7990 years per synonymous mutation over the mitochondrial genome. However, the distribution of substitutions at synonymous sites in human data significantly departs from a model assuming a single rate parameter and implies at least 3 different subclasses of sites. Neutral model with 3 synonymous substitution rates can explain most, if not all, of the apparent molecular clock difference between the intra- and interspecies levels. Our findings imply the sluggishness of purifying selection in removing the slightly deleterious mutations from the human as well as the Neandertal and chimpanzee populations. However, for humans, the weakness of purifying selection has been further exacerbated by the population expansions associated with the out-of Africa migration and the end of the Last Ice Age.

Link

9 comments:

Maju said...

I'm still reading it but so far I have two opinions:

1. It is only normal to expect that the number of mutations should grow as population grows (more people = more novel mutations necesarily). However that does not mean these mutations will necesarily survive or will become frequent enough as to be noticeable in the usual samples.

2. Notice (fig. 3) that the central referential ages for Sapiens-Neanderthal and Homo-Pan divergences are in the shortest imaginable ranges (440 Ky and 6.5 My respectively) and are not something I can admit. I have already argued that the Pan-Homo divergence must be older (at least 8 My maybe as much as 10 My) and also that the Neanderthal-Sapiens divergence must be of c. 900 Ky. (I don't believe in an universal "Homo heidelbergensis" but in different Erectus-derived lineages for the two last human species: H. heidelbergensis in Europe for Neanderthal and H. rhodesiensis in Africa for H. sapiens).

Hence, and taking the Neanderthal reference as safer, I'd argue that all the age estimates of this paper (assuming that everything else is correct) should be doubled in age (i.e. twice older).

Maju said...

Correction: double ages sound good for the Sapiens-Neanderthal divergence but for the Pan-Homo one the difference is not as extreme but rather 23-54% older. I discussed the Pan-Homo divergence issue here.

Maju said...

Also I'd like to know what do people think of the fact that the two apparent signatures of expansion in this study (fig. 2) are dated by the authors to (1) c. 55 Kya and (2) c. 15 Kya, suggesting that the OoA expansion happened much later than their own age estimates for mtDNA M and N (77 Kya and 68 Kya respectively) and also quite some time before Neolithic (with a decrease of the expansion signature in fact after 10 Kya, i.e. after Neolithic).

I make little sense of this and makes again be extremely wary of pure molecular clock estimates without proper hard data to confirm these ages.

Finally, has anyone read the briefly mentioned Endicott et al. 2009, which argued that mtDNA H and U are as old as 75 Kya?

Maju said...

And finally (and sorry for posting so fragmentarily), has anybody opinions on table S1, with age estimates for some R sublineages that are clearly older than their age estimate for their ancestor N (R30 and R31 yield ages of 75 and 78 Kya respectively)?

Other oddly old age estimates are in that table are U8 (56 Kya - i.e. long before the colonization of West Eurasia) and H1 (31 Kya - much older than the ages of other H subclades).

terryt said...

"However that does not mean these mutations will necesarily survive or will become frequent enough as to be noticeable in the usual samples".

I think the argument here is that they will survive if there is very little selection. As Dienekes wrote, 'as populations expand, even slightly deleterious alleles (and even non-deleterious ones) have a higher chance of surviving in a population'. And selection automatically reduces the number of mutations even if the selection is not actually acting on those mutations.

"I don't believe in an universal 'Homo heidelbergensis' but in different Erectus-derived lineages for the two last human species: H. heidelbergensis in Europe for Neanderthal and H. rhodesiensis in Africa for H. sapiens"

I don't think you're justified in that belief. I can sort of concede your chimp/human split could be correct. But then that still wouldn't place the Neanderthal/modern human split as long ago as you would like.

"I'd like to know what do people think of the fact that the two apparent signatures of expansion in this study (fig. 2) are dated by the authors to (1) c. 55 Kya and (2) c. 15 Kya, suggesting that the OoA expansion happened much later than their own age estimates for mtDNA M and N (77 Kya and 68 Kya respectively) and also quite some time before Neolithic (with a decrease of the expansion signature in fact after 10 Kya, i.e. after Neolithic)".

For a start I strongly suspect that their narrowing their idea to just the two postulated expansions followed by selection is a considerable underestimate, especially if we look at the combined haplogroups, and bound to lead to inconsistencies. If there have been several periods of what the authors call 'purifying selection' in various populations even the possibility of comparative dating between different haplogroups dissappears. It would also make precise dating of any OoA movement impossible.

Maju said...

... they will survive if there is very little selection.

No because you also have to consider drift. Drift probably annihilates more mutations than selection, even some favorable ones.

As Dienekes wrote, 'as populations expand, even slightly deleterious alleles (and even non-deleterious ones) have a higher chance of surviving in a population'. And selection automatically reduces the number of mutations even if the selection is not actually acting on those mutations.

The greater the population the weaker both the drift and the selection. In fact weak selection is not too different from drift, just that the adaptative alleles have slightly better chances of survival. But chances are not any guarantees, just slightly better odds.

I don't think you're justified in that belief. I can sort of concede your chimp/human split could be correct. But then that still wouldn't place the Neanderthal/modern human split as long ago as you would like.

There are other people who think that way: the existence of a single parent species for Neanderthals and Sapiens after 900 Kya is not justified on archaeological grounds.

The correction for the Pan-Homo split may not justify that extension from the viewpoint of the molecular clock complicated and controversial theory but that means little in comparison. Specially as the various hominid branches could well have also suffered from various accelerations and decelerations as population figures changed in all that time, something not considered here or anywhere that I know of.

If there have been several periods of what the authors call 'purifying selection' in various populations even the possibility of comparative dating between different haplogroups dissappears. It would also make precise dating of any OoA movement impossible.

These may be wise words.

Gioiello said...

What I have supported about the YDNA, i.e. the calculation of Nordtvedt, Klyosov etc. (I don’t take in consideration Vizachero) is wrong, because it doesn’t take in consideration the mutations happened around the modal, it is also worth for mtDNA. The youngest clades, like that of my daughter K1c1*, have less mutations than my K1a1b1, because they arose from an old K1 purified during the time from deleterious mutations. Then everything has been said about the extinction and the non continuity of European ancient mtDNA from Paleolithic to Neolithic is wrong, because only a few clades from that time have survived, were purified and expanded.

Gioiello said...

I thank you for your aid and of course you'll be able to see my results also before I submit them to GenBank. My reasoning was this: the paper in object says that K1c1* (my daughter), mutations 9093-11377, was born from a K1c 3100 YBP. My K1a1b1 (11914) was born 22500 YBP… For this I had thought that a K1 had survived for a long time… but K1c1 was born from a previous K1c and K1a1b1 from previous subclades. Then the dates of the paper are misleading. We must presuppose that the subhaplogroups a, b, c were born from K1 at about the same time, and my reasoning then was wrong. In Y we are now discovering some SNPs that differentiate, for instance, R-P312 before the subclades. For mtDNA we haven’t the same, except perhaps heteroplasmy, but anyway after so much time any trace is deleted. But like in Y, probably not all subclades accumulate during the time the same number of mutations, and this is a problem to explain.
For the Y there is something not convincing in the calculation of the TMRCA. I wrote this to Nordtvedt:
“Ken, I know better my R1b1b2.
DYS426 is a very slow mutating marker.
R1b1* had 12
R1b1b2 (L23-) had 11
R1b1b2/L23+ (mine) had and has 12
R1b1b2/L51+ had and has 13
All subclades have 12
From R1b1* to R1b1b2a1b may have passed 40,000 years and not a few thousands. Probably they have passed less than 40,000, but certainly not about 6,000 as Vizachero pretends and hopes.
In the meanwhile faster mutating markers have changed many times around the modal and perhaps now are the same of the origin, except someone: see DYS385: R1b1b2 14-11, R1a1 11-14 etc.
This is my thought (and my hope if you want)”.

eurologist said...

the existence of a single parent species for Neanderthals and Sapiens after 900 Kya is not justified on archaeological grounds.

Maju,

Of course the problem is that there are few remains in those regions to start with. Still, heidelbergensis had about half a million years to enter North Africa, during varied climatic periods. During some such phases, it would have been able to cross into more southern regions through the central Sahara and also via the Nile valley.

I wouldn't propose that such hypothetical intruders were the progenitors of AMHs. But at least until Neanderthal genetic comparisons have become more refined, we can't exclude some (perhaps even significant) mixing, either. If the Neanderthal studies turn out to be sufficiently reliable, it may be possible to identify "blocks" of material that were shared ~400,000 years ago, if there are any.

Oh, and many thanks for all your other insightful comments with regard to this paper.