PDA

View Full Version : Mitochondrial haplogroup origin dating & placing, is it actually that reliable?



PerceptionDeception
02-15-2018, 12:24 PM
I am reading "Blood of the Isles" by Bryan Sykes written in 2006 and he is beginning to explain how mitochondrial haplogroups are categorized and the time of origin determined.

In haplogroup T, he states that there are only 2 "signature" mutations among a couple of other later mutations that can be found scattered around Europe. The way he describes dating when the Clan Mother was around doesn't seem to make much sense to me. He states that typically there is one mutation in the mitochondria every 20'000 years, and since the average amount of mutations that have occurred is 0.85, haplogroup T came about roughly 17'000 years ago.

Is there more supporting evidence for when haplogroups originated or is that it? It seems like pseudoscience at best to me because a) How do we know that 20'000 years is the average rate of mutation? and b) Is there not a chance that the haplogroup is 60'000 years old and that by luck no mutations have taken place?

I'm interested in further evidence and methods used to date haplogroups.

In addition, I feel the technique he is discussing to estimated the place of the haplogroups origins is speculative. He states that since there are a greater amount of additional mutations in certain parts of the haplogroup territory than others, these must be the origin point because the mitochondria has had more time to mutate in these old places. I don't believe that is conclusive proof because the mutation could have taken place anywhere. This doesn't make any sense to me as a method for determining its geographic location.

I'd really appreciate help on explaining this further to me.

Thanks in advance.

msmarjoribanks
02-15-2018, 03:17 PM
Blood of the Isles is outdated and a lot of the dating is off, as I understand it. A lot of the haplogroups that he identifies as hunter-gatherer likely came with early farmers (K, for example -- K1a and K1b plus K2a, if memory serves, whereas K2b seems to be if anything associated with the spread of R1a). I'm just focusing on K because I know the most about it.

Jean Manco has a great site with tables of the testing results for lots of ancient DNA: http://www.ancestraljourneys.org/ancientdna.shtml

We need more detailed analysis (saying U or K or H doesn't tell us much because mutations are so slow), and also more detailed analysis of patterns within current populations, IMO.

Good source to start with now for what we know is Eupedia (https://www.eupedia.com/europe/origins_haplogroups_europe.shtml) and http://haplogroup.org/, although obviously there are lots of papers to drill down more specifically.

C J Wyatt III
02-15-2018, 04:30 PM
....Is there more supporting evidence for when haplogroups originated or is that it? It seems like pseudoscience at best to me because a) How do we know that 20'000 years is the average rate of mutation? and b) Is there not a chance that the haplogroup is 60'000 years old and that by luck no mutations have taken place?

I'm interested in further evidence and methods used to date haplogroups.

In addition, I feel the technique he is discussing to estimated the place of the haplogroups origins is speculative. He states that since there are a greater amount of additional mutations in certain parts of the haplogroup territory than others, these must be the origin point because the mitochondria has had more time to mutate in these old places. I don't believe that is conclusive proof because the mutation could have taken place anywhere. This doesn't make any sense to me as a method for determining its geographic location....


I believe you are onto something. I am known for out of the mainstream views, but I think both mutations and back mutations occur at a much higher rate than currently believed. A unbroken line of a particular mtDNA subclade for thousands of years then becomes very unlikely. Anyhow, I am not leaning on anyone anymore to take the expensive FMS until more is known, though it is nice to know the upper level approximations you get from certain services.

Jack Wyatt

PerceptionDeception
02-15-2018, 04:42 PM
I have noticed Blood of the Isles is outdated in many things it says. It's the first book on genetics I've ever bought and I know in future to buy with caution as genetics seems to be such a rapidly changing field.

Megalophias
02-15-2018, 05:29 PM
In haplogroup T, he states that there are only 2 "signature" mutations among a couple of other later mutations that can be found scattered around Europe. The way he describes dating when the Clan Mother was around doesn't seem to make much sense to me. He states that typically there is one mutation in the mitochondria every 20'000 years, and since the average amount of mutations that have occurred is 0.85, haplogroup T came about roughly 17'000 years ago.
It sounds like he is talking about only HVS-I (hyper-variable sequence 1), which covers positions 16024 to 16383 (out of 16569) in the standard reference sequence of human mitochondrial DNA. This is only about 2% of the mitochondrial chromosome, but much more prone to mutate than most of the chromosome. So in order to save money people will often sequence only HVS-I, or that and a few other small sections, rather than the whole chromosome.

Haplogroup T has only 1 HVS-I mutation defining it (C16294T, that is a transition of cytosine to thymine at position 16294). However, it has 9 other mutations in the coding region (the main part, full of important genes).


Is there more supporting evidence for when haplogroups originated or is that it? It seems like pseudoscience at best to me because a) How do we know that 20'000 years is the average rate of mutation?
We know what the average rate of mutation is 1) by looking at loads of modern parents and children to see how many mutations have happened per generation on average and 2) by counting the average number of mutations that have accumulated over a known period of time. In early research this was based on a lot of inference and assumptions, trying to associate haplogroups with fossil or archaeological evidence of known (or guessed) date; but now we have radiocarbon-dated ancient DNA and we can actually count how many more mutations on average people have now than they did 45 000, 40 000, 30 000 years ago. That allows us to make a fairly solid estimate of the mutation rate. Mind you it is still possible and in fact very likely that the average mutation rate can change in different times and places, but probably not by very much.


and b) Is there not a chance that the haplogroup is 60' 000 years old and that by luck no mutations have taken place?
Sure. For instance the human reference sequence, which belong to haplogroup H2a1a1, has not experienced an HVS-I mutation in its lineage since the birth of haplogroup R itself around 50 000 years ago. This is very unusual but not wildly unlikely. But, this is only one branch of haplogroup H, which is only one branch of R0, which is only one branch of R, each of these levels having many many branches. We look at the average number of mutations accumulated by many descendants of one haplogroup; one line (subhaplogroup) may have had extaordinarily few or extraordinarily many mutations, but the chance of many lines doing so at once is very remote. Also, we can look at the whole sequence rather than just the HVS-I; in the whole sequence mutations come at an average of about one every 2500-3000 years.

If you look at the actual estimates for mitochondrial haplogroup TMRCAs given in scientific studies you will see that they come with large error bars; it won't be "25 000 years" but more likely "16 000-34 000 years" or something like that.


He states that since there are a greater amount of additional mutations in certain parts of the haplogroup territory than others, these must be the origin point because the mitochondria has had more time to mutate in these old places. I don't believe that is conclusive proof because the mutation could have taken place anywhere. This doesn't make any sense to me as a method for determining its geographic location
Yes, it is speculative and not conclusive proof of anything. Basically the idea is that if a haplogroup developed in one area and then spread to another area through people migrating, then probably only a subset of the haplogroup's subclades will be represented in the new area, since usually only a subset of a population will migrate to another place. So the diversity of the haplogroup is expected to be greatest in the home area. This is the centre of diversity argument and has been applied to all kinds of things, like crop domestication and the homelands of languages. However it is really only a rule of thumb or good default hypothesis, since the actual history could produce a misleading distribution.

These estimates are not pseudoscience, but should be taken with due caution.

Onur Dincer
02-15-2018, 06:07 PM
The thread has been moved from the General section to the Inquiries Corner section.

Saetro
02-15-2018, 08:23 PM
I have noticed Blood of the Isles is outdated in many things it says. It's the first book on genetics I've ever bought and I know in future to buy with caution as genetics seems to be such a rapidly changing field.

It is engaging, inspiring and easy to read.
There are copies around in my public library system, so I can recommend it to people wondering about the field.
But the field was changing rapidly and it represents an early position.


Megalophias
It sounds like he is talking about only HVS-I (hyper-variable sequence 1)
Yes, and the YDNA involves only 7-10/11 STR markers.

The science has moved on.
As CJ Wyatt III mentioned, FMS is expensive, but I do notice more people (mostly women) doing it once they feel comfortable with aDNA.
Whatever the AVERAGE mutation rate, every individual is different, so you may have two mutations in one generation or none for 1000+ years.
This is very much outside ordinary experience and often does not help find cousins directly.
I know of mtDNA being used to prove a suspected close relationship.
Few if any of my friends have found cousins for whom documentary connections can be found.
But they have been able to narrow the geographical area their all-female line ancestors came from.
As have I, and this has been very helpful in excluding some possible links along that line.

msmarjoribanks
02-16-2018, 04:30 AM
It is engaging, inspiring and easy to read.
There are copies around in my public library system, so I can recommend it to people wondering about the field.
But the field was changing rapidly and it represents an early position.

Agree with this. It's actually the book that got me interested in DNA testing, that and his one about the USA.

msmarjoribanks
02-16-2018, 04:42 AM
Whatever the AVERAGE mutation rate, every individual is different, so you may have two mutations in one generation or none for 1000+ years.

I've mentioned this before (so apologies for being repetitive!), but I've seen results similar to both of these.

I have 1 exact match full test and 1 one mutation removed test. The one mutation removed (she has one mutation I don't) is a Family Finder match, so her mutation happened after our connection (which was likely in the 1800s in the US, although we haven't pinpointed it yet, and it has been helpful in confirming some suspicions about where I thought the line was located before my current brick wall).

On the other hand, my dad has many full matches and many more one mutation off matches. His line traces back through 1630s immigrant to Massachusetts from Suffolk, and her family is traced back to a woman in the mid 1500s (we have good documentation of this line, but it was nice to have the confirmation through the matches). Several of his full matches and a couple of the one mutations trace back to this same line, matching up at different places. Others, though, do not. Among the full matches is someone whose ancestry goes back to Norway in the 1700s, and another who traces back to a woman who immigrated to the US in the 1800s from Germany. It seems like those must be OLD connections, and yet no mutations.

ThirdTerm
02-16-2018, 08:57 PM
The Seven Daughters of Eve (2001) by Bryan Sykes is a good introductory book on human mitochondrial DNA, while methods of studying mitochondrial genomes have advanced greatly since the 2000s. When he wrote these bestselling popular science books on human genetics, peer-reviewed scientific papers on human mitochondrial DNA were hardly available. There are limitations in calculating the mtDNA substitution rate based on fossil calibration or archaeological evidence. To circumvent the problem, Fu et al. (2013) calculated the human mtDNA substitution rate directly from complete mitochondrial genomes of ten ancient modern humans, for which reliable radiocarbon dates are available.



An alternative approach for obtaining greater precision in measuring substitution rates is through the analysis of genetic data from ancient samples for which reliable radiocarbon dates are available. Ancient humans are well suited to provide calibration points for the human mitochondrial molecular clock: reliable radiocarbon dates are available for many specimens; hence, the number of substitutions that have accumulated among lineages can be directly translated into the number of substitutions per site per year. Branch shortening—the effect of fewer nucleotide substitutions on ancient branches in a phylogenetic tree as compared to modern—is commonly observed in phylogenetic studies of ancient humans [10]. The observed branch shortening reflects the comparatively shorter time since the common ancestor for the ancient human as compared to the present-day individual: a present-day lineage has had more time to accumulate nucleotide changes. If we know the age of the ancient sample (e.g., from a carbon date), we can thus infer the mutation rate necessary to produce the observed degree of branch shortening.

Here we use the complete or nearly complete mitochondrial genomes from ten ancient modern humans for which reliable radiocarbon dates are available to calculate the human mtDNA substitution rate directly. This strategy circumvents the limitations imposed by the use of indirect measures of substitution rates such as those obtained via fossil calibration. The samples used in this analysis span 40,000 years of human history and originate from Europe and Eastern Asia. We use our substitution rate to estimate the dates of major human evolutionary events in the last 200,000 years. Of particular note, our rate suggests an upper bound on the split between non-Africans and sub-Saharan Africans of less than 95 kya. Even though this estimate is a conservative upper bound because the genetic divergence at any particular locus in the genome is by definition older than the population divergence time, this date, is on the extreme lower end of population split times estimated from de novo substitution rates for nuclear DNA [1].

https://ars.els-cdn.com/content/image/1-s2.0-S0960982213002157-gr1.jpg

We were able to reconstruct three complete and six nearly complete mitochondrial genomes from ancient human remains that were found in Europe and Eastern Asia and span 40,000 years of human history. All Paleolithic and Mesolithic European samples belong to mtDNA hg U, as was previously suggested for pre-Neolithic Europeans [15]. Two of the three individuals from the Dolni Vestonice triple burial associated with the pre-ice age Gravettian culture, namely, 14 and 15, show identical mtDNAs, suggesting a maternal relationship. Furthermore, both individuals display a mitochondrial sequence that falls basal in a phylogenetic tree compared to the post-ice age hunter-gatherer samples from Italy and central Europe, as well as the contemporary mtDNA hg U5 (Figure 1). It has been argued that hg U5 is the most ancient subhaplogroup of the U lineage, originating among the first early modern humans in Europe [18]. Our results support this hypothesis because we find that the two Dolni Vestonice individuals radiocarbon dated to 31.5 kya carry a type of mtDNA that is as yet uncharacterized, sits close to the root of hg U, and carries two mutations that are specific to hg U5. With our recalibrated molecular clock, we date the age of the U5 branch to approximately 30 kya, thus predating the LGM. Because the majority of late Paleolithic and Mesolithic mtDNAs analyzed to date fall on one of the branches of U5 (see also [15]), our data provide some support for maternal genetic continuity between the pre- and post-ice age European hunter-gatherers from the time of first settlement to the onset of the Neolithic. U4, another hg commonly found in Mesolithic hunter-gatherers [15], has so far not been sequenced in a Paleolithic individual, and we find hgs U8 and U2 in pre-LGM individuals but not in later hunter-gatherers. At present, the genetic data on Upper Paleolithic, and especially pre-ice age, populations are too sparse to comment on whether or not this is representative of a change in the genetic structure of the population, perhaps caused by a bottleneck during the LGM and a subsequent repopulation from glacial refugia.

https://www.sciencedirect.com/science/article/pii/S0960982213002157

Saetro
02-18-2018, 01:20 AM
The Seven Daughters of Eve (2001) by Bryan Sykes is a good introductory book on human mitochondrial DNA, while methods of studying mitochondrial genomes have advanced greatly since the 2000s. When he wrote these bestselling popular science books on human genetics, peer-reviewed scientific papers on human mitochondrial DNA were hardly available. There are limitations in calculating the mtDNA substitution rate based on fossil calibration or archaeological evidence. To circumvent the problem, Fu et al. (2013) calculated the human mtDNA substitution rate directly from complete mitochondrial genomes of ten ancient modern humans, for which reliable radiocarbon dates are available.

"Blood of the Isles" reads now like dated science.
I also read "7 daughters of Eve" when it came out, but now it reads like fiction: maybe one of Kipling's Just So stories.
Can still be inspiring these days, but not for everyone.

Fu's paper is wonderful.
Just after it came out there were comments that these 10 samples did not represent the branch that they knew and so on.
(Probably still tucked away in this forum. Look for old posts.)
So Fu (2013) is still not regarded as the last word.

The important point is that averages represent groups, but we each test as individuals who can be widely deviant from the average.

Judith
02-18-2018, 01:12 PM
Blood of the Isles was one of the first books I read too. An easy read and inspired me to learn more. The original scientific papers are hard work and you may only understand the introduction and discussion the first time through but I have more on each reading. Even the papers are overtaken virtually every year this is a field which is developing fast.
Use google scholar to find the papers and the use cited by to find papers which build on it.
Depending which haplogroup you are looking for you will find lots or just some information. Mtdna H is the most difficult because authors just report H and all H clades are NOT the same.
Also useful are ftdna projects where you can read without joining and Eupedia has been mentioned. Treat all with a pinch of salt. There are more blogs with opinion than facts.

TuaMan
03-19-2018, 01:33 PM
These estimates are not pseudoscience, but should be taken with due caution.

Hey Mega, question for you (or anyone else who knows, feel free to chime in as well) -

In this paper (http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0129839), if you look at tables 2 and 3 with the mtDNA coalescence ages, you notice there's a big discrepancy for many mtdna between the "observed age" and the "expected age." For example if you look at mtdna X, the authors calculated an observed age of 31.9 kya, versus an expected age of 66.6 kya. Why the massive difference between the two? I tried skimming the paper to find out how they calculated observed vs expected and if they mention anything about reconciling the two but I couldn't find anything.

Megalophias
03-19-2018, 02:22 PM
In this paper (http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0129839), if you look at tables 2 and 3 with the mtDNA coalescence ages, you notice there's a big discrepancy for many mtdna between the "observed age" and the "expected age." For example if you look at mtdna X, the authors calculated an observed age of 31.9 kya, versus an expected age of 66.6 kya. Why the massive difference between the two? I tried skimming the paper to find out how they calculated observed vs expected and if they mention anything about reconciling the two but I couldn't find anything.
In this paper they are trying to figure out where haplogroup N spread from, and the idea is that the haplogroup N subclades should be older near where the expansion began, and younger as they get further away. The model they are testing assumes the expansion began in East Africa (Djibouti) at the age of L3, and ended in Australia at the age of haplogroup S, so the further away from Djibouti the haplogroup is centred, the younger the expected age will be. Since haplogroup X is found in West Eurasia and Africa, it is close to the presumed starting point, and so according to this model its expected age is very old.

So the expected age is just a feature of this model calculated based on distance from Africa, it is not anything applicable outside of this paper. The observed age is the actual genetic TMRCA of the haplogroup. There is a big gap between them because haplogroup X didn't originate immediately after Out-of-Africa, the model is a poor fit.

TuaMan
03-19-2018, 02:39 PM
The observed age is the actual genetic TMRCA of the haplogroup. There is a big gap between them because haplogroup X didn't originate immediately after Out-of-Africa, the model is a poor fit.

Hmm, I get that TMRCA of 30,000 kya means its most recent coalescence is way too young for OOA, but isn't X a primary branch of N, which is at least 60,000 years old? So couldn't X also have formed nearer to 60,000 years ago shortly after the coalescence of N, but just went through a big bottleneck until 30,000 years ago when it began its first real expansion?

Megalophias
03-19-2018, 03:34 PM
Yes, that's right. X would have formed (by definition) when N coalesced, since it is a primary branch of N, and then must have had a long period without growth.

The idea is that the *average* age of haplogroup expansion (not just considering one case) should be greater in the areas first colonized by modern humans. This does not necessarily make sense, though. For instance West Asia and North Africa were quite arid, so they could have had smaller population sizes and more bottlenecks during the Ice Age than relatively warm and moist South and Southeast Asia, regardless of where modern humans reached first.