mtDNA K=UHG?

I've got a really strong feeling that the ancestry related to Mesolithic Europeans ('UHG') in Neolithic Anatolia derives from a population which mostly carried mHG K. These mHG K rich people (UHG) would have been the Near Eastern/Eastern European cousins of mHG U5 rich Mesolithic Europeans (WHG). I think they lived somewhere within this range that I guess you can call the 'East Mediterranean.'

Click  HERE  for a PDF version that is easier to read.***   

I've scratched off a big chunk of Europe from my list of possible homelands for K1 and K1a because the evidence proves a large chunk of K1 in Europe arrived from Anatolia in the Neolithic(See the map below). By and large K1a in (Western?) Europe falls under a handful of subclades dated to the early Neolithic: K1a4a1(9ky), K1a2a(8ky), K1a3a(10ky), K1a1b1(8ky). All of those clades have been found in Neolithic farmers from Spain and Britain but not in any Mesolithic European foragers.

Click  HERE  for a PDF version that is easier to read.*** 

Here are phylogenetic trees I made for K1. The trees include age estimates I get for K1 clades using Soares 2009's age estimator.

K1a tree. *Blank boxes are for rare K1a clades. I only show the name of popular K1a clades.

Click  HERE  for a PDF version that is easier to read.*** 



Other K1 tree.

Click  HERE  for a PDF version that is easier to read.*** 

In the Middle East, and possibly some parts of Europe, K1a has few links to K1a in (Western?) Europe. The only K1a clade common in both Europe and the Middle East is K1a4 (13ky). This means most European (ultimately Anatolian) and Middle Eastern K1a clades went their separate ways way back in the Early Mesolithic.

mHG U5b1 (19ky) shows this same pattern. Most U5b1 clades in Spain and Eastern Europe went their separate ways back in the Mesolithic/Late Paleolithic. Ancient mtDNA confirms U5b1 (19ky) clades had already expanded and gained popularity by the Mesolithic. The same may turn out to be true for K1a (17ky).

Mesolithic K1a didn't have the same geographic extent it has today. I think K1 and K1a were expanding, diversifying into new subclades, only within some populations of the 'East Mediterranean.' Anatolia Neolithic's mysterious 'UHG' ancestor was definitely one of those populations.

There's actually already documentation of K1 in the Mesolithic 'East Mediterranean'. The first two mtDNA samples from Mesolithic Greece belong to the same rare or extinct K1c* clade (See here) and 26% (11 out of 42) of mtDNA samples from Mesolithic Serbia/Romania belong to now rare or extinct forms of K1(K1*=6, K1c*=3, K1a*=1, K1f*=1).

Watch when someone sequences DNA from Mesolithic Anatolia. I won't be surprised at all if the results are a WHG-like population with 50% mHG K1 that fits well as Anatolia Neolithic's 'UHG' ancestor. 

The evidence says one thing, they say another

While searching for new mtDNA data on google today I came across gold mine: Olivieri 2013 et al. This study sequenced  many mitogenomes belonging to mHGs W and N1a1b, two widespread and poorly understood lineages.

Though I do appreciate the amazing work the people who took part in this study did, the conclusion they made about N1a1b is complete bonkers. Their conclusion isn't bonkers because they aren't aware of other data which disproves their conclusions, it's bonkers because their own data dis agrees with their conclusions.

The conclusion/assumption they came up with is that haplogroups N1a1b  took part in repopulating Europe after the Ice age('Last Glacial Maximum') about 15,000-12,000 years ago. But if they looked at the their own age estimates for mHG N1a1b clades they would have seen this isn't very probable.

A large majority of European N1a1b is I1a1, I3a, and I2. Look at the age estimates Olivieri 2013 got for those clades.

I1a1; 4,900yp
I2; 6,800yp
I3a: 7,400yp

The LGM ended 12,000yp, long before any of these clades are estimated to have began. How did Olivieri 2013 not see this? Ancient mtDNA shows I1a1, I2, and I3a likely spread into Europe from the Russian Steppe about 4,000-5,000 years ago. The oldest example of I3a is in Poltvaka. Plus Andronovo, Bell Beaker, and Unetice had a really high frequency of these mHGs.

I use Soares 2009's age estimator to get estimates for mtDNA haplogroups and it gets similar age estimates for these N1a1b clades as Olivieri 2013 did.

Soares 2009's estimator gives similar age estimates for all the major European mtDNA haplogroups. So what this mans is a large majority of European mtDNA belongs to mHGs that, according to Soares 2009's age estimator, arose less than 10,000 years ago. That's consistent with what ancient mtDNA shows; European mtDNA mostly derives from two migrations that occurred between 5,000yp and 10,000yp and originated from relatively small regions with *relatively* low mtDNA diversity.

This isn't an isolated incident, Olivieri 2013 is just one example of a host of mtDNA studies which make the same naive assumption about European origins. They all assume the current European gene pool derives from southern refugiums that repopulated Europe after the Ice age roughly 15,000 years ago.  Oh yeah and they also think modern day Basque are basically fossils from Ice age Europe.

No matter what the data actually indicates, they come to the same ole conclusions in their studies. Well, now that ancient DNA has thoroughly disproven these assumptions, hopefully mtDNA studies about the peopling of Europe will begin to explore more possibilities.

***
They aren't horrible people for making these assumptions. It's human nature we all make assumptions like this. 5-10 years ago, it appeared the evidence suggested the European gene pool formed during the LGM. That's what all the smartest experts though. I also tend to make assumptions based on what the most popular current theories are. I'm sure many current theories will be proven wrong in the future. 

I can understand why a geneticists in 2013 would think modern European origins can be explained mostly by LGM migrations but geneticists who think this in 2017 have no excuse. Commercial DNA testing companies like 23andme and FTDNA still preach these old disproven theories as fact to their customers. There's no excuse for this, data that disproves these theories has been around for several years now. 

Out of Iran

*At the bottom of this post I have listed all of the mHGs I think *may* have expanded out of somewhere near Iran. I used the mutation rate created by Soares 2009 to create age estimates.

I’m begging to notice an interesting trend in West Eurasian mtDNA that reinforces recent findings from ancient genomes. Quite a few mHGs look like they expanded out of Iran and surrounds mostly after 10,000 years ago. That time frame coincides with the out of Iran/Caucasus into Europe and SW Asia migrations documented by ancient genomes.

Initial reports about genomes from Neolithic Iran presented them as completely unrelated to other early Neolithic farmers, see here. But I think the differences between Iran Neolithic and other ancient farmers has been exaggerated. mtDNA data suggests all early Middle Eastern farmers shared lots of mtDNA.

Based on modern mtDNA and some ancient mtDNA, I’m fairly confident Iran Neolithic had loads of mHG J1, W, N1a, and HV just as Anatolia Neolithic did. Yes, farming was started in Iran and Anatolia by different populations but early farmers in both regions certainly had common Paleolithic hunter gatherer ancestors. They were probably more related than what is currently thought. The genetic landscape of the Paleolithic/Neolithic Middle East is definitely pretty complex and won’t be fully understood till we get more ancient genomes.

R2’JT
mHG, Most Common in, Age Estimate
J1d1a1, Levant and Arabia, 6ky
J1b1a1, Europe and South Asia(found in Catacomb and Corded Ware), 6ky
J1b1b1, entire Middle East and South Asia, 12ky
T1a1, Europe and South Asia(found in Yamnaya, Srubnaya, BA Europe), 7ky
T2i2, South Asia, n/a
T2d1a, South Asia, n/a
R2, most of Middle East and South Asia(found in Iran Neolithic), n/a

HV
mHG, Most Common in, Age Estimate
H13a1a, Europe(Found in Yamnaya, BA Europe), 8ky
H13a2a, Iran and South Asia, 11ky
H15, West Eurasia, n/a
H6a1, Europe(Found in Yamnaya, BA Europe), 8ky
HV2, Iran and South Asia(found in Iran Neolithic), n/a
HV12b1, Iran, Caucasus, South Asia, n/a

Other
mHG, Most Common in, Age Estimate
U1a1, Middle East and South Asia, 15ky
U7, Middle East and South Asia(found in Iran Chalcolithic), n/a
U2a, U2b, U2c; South Asia, n/a
I1, West Eurasia(found in BA Europe), n/a
I2, Europe, n/a
I3, Europe(found in Poltvaka and BA Europe), n/a
W6, West Eurasia(found in Yamnaya and BA Europe), 9ky
W3a1, South Asia and Europe(found in Yamnaya and Corded Ware), n/a
W4a, West Eurasia, n/a

The age of ULTIMATEness!

A new age for mtDNA Atlas has begun; The age ULTIMATEness!! This new age is named after ‘ULTIMATE’ which is a spreadsheet of mine that currently contains 33,185 mtDNA samples. I won’t share ULTIMATE with anyone.

ULTIMATE allows me to look at mtDNA data at angles I could never do before. Thanks to ULTIMATE my knowledge of West Eurasian mtDNA has quadrupled over the past month. Stay tuned for some interesting posts in the second half of 2017.

ULTIMATE can show me the distribution of any mtDNA haplogroup or mtDNA haplotype in the blink of an eye.

I used ULTIMATE to do a haplotype analysis of about 20,000 samples from Europe and the Middle East. What I found is pretty interesting. Most of the stuff I've learned I'm going to keep top secret for a while. I have lots of new theories on the ancient origin of mHGs that I will post about soon. I think many of you will find them interesting.

Haplotype sharing correlates with geography pretty well. Because of my haplotype analysis of West Eurasia I can classify about 10% of the mtDNA in every population in West Eurasia to a specific sub region such as Iberia or Balkans or Eastern Europe.

Here’s an introduction to what will be the new norm on this blog.

Haplotype sharing results.

Poland.

Macedonia.

Lebanon.

16 unique Balkan haplotypes

I’ve discovered 16 haplotypes unique to the Balkans. Combined they make up about 5% of total Balkan mtDNA and a whopping 10% of the mtDNA in Eastern Herzegovinian speakers (Bosnia, Serbia, Kosovar, Croatia). Most of these haplotypes have a representative in four or more Balkan populations. That confirms all of these haplotypes are widespread in the Balkans but pretty much never found anywhere else in the world and can therefore be called ‘Balkan haplotypes.’

This map gives an idea where the haplotypes are found….

They are not as frequent in all Balkan populations. Greece and Slovenia only have 1%. The haplotypes seem to specifically be centered around the relatively small range of former Yugoslavia.

It’s tempting to then argue that they are 'Yugoslavian haplotypes'. But that wouldn’t be able to explain why Albanians and Bulgarians have almost as much as former Yugoslavia.

The haplotypes did pop up in some non Balkan populations; Estonia, Ukraine, Iraq, and Southern Italy. Southern Italy has the most. Though Southern Italy is separated by ocean, it almost kisses the Balkan peninsula, so it would make sense that Southern Italy has a lot of Balkan ancestry.

The presence of four of these haplotypes in Southern Italy indicates they are pre-Slavic because Southern Italy doesn't seem to have any Slavic ancestry.

Balkan haplotypes.

H1c 16042 16288
H1a 16258c
H13a2c1
H 16189! 153 204
HV2(a) 16243 16261
R0a2c 16168 16265
U5b1 16293 16186 16189! 16192!
U5b1a 204 207 235
U4c1a 16362 16051G
U2e(1) 16209
U1a(2) 16179t 16129!
W1c'i 16193t 152!
W 16172 16231 [199]
N1b1 16244A
N1b1 16241g 245c

U5a1d2b, an Eastern U5a1 branch.

Haplogroup U5a1d2 has two branches; U5a1d2a and U5a1d2b. U5a1d2a is more or less exclusively European but U5a1d2b has a much more eastern distribution. U5a1d2b is more or less exclusively found in non Indo Europeans from far eastern Europe and Siberia. It peaks in the Mari of Russia and the Tubular of the Atlai region in Siberia.

Ancient DNA supports the idea U5a1d2b is sometype of far eastern/north Eurasian U5a1 branch. It was relatively common in Siberian Afanasievo, Andronovo, and later in Siberian Sycthians who lived in the same land as modern Tubular. U5a1d2b has also popped up in Comb Ceramic twice, Yamnaya once, and Sarmatians once.

U5a1d2b's strong presence in both Siberian Afanasievo and Siberian Scythian indicates that Scythians had some Afanasievo ancestry.

U5a1d2b's presence in Saami might be able to be explained by Comb Ceramic, considering Sycthians never lived near Scandinavia. And its strong presence in both Volga Russia and Siberia definitely looks like a legacy of the Sycthians.

LBK-like mtDNA in Lengyel Poland (Cylenski 2017)

In March, Cylenski 2017 published one mito genome from Mesolithic Poland, one from LBK Poland, and three from Lengyel Poland. This study went unnoticed by DNA forums. I discovered it while searching for studies on modern mtDNA.

Here are the results from the paper....

Mesolithic, 5644-5374 cal BC, Jan1, U5b1b1
LBK culture, ----, Sam1, N1a1a1a
Lengyel culture, ----, Kz6, N1a1a1a3
Lengyel culture, ----, NHp1, H5*
Lengyel culture, 4255–4145 cal BC, R18_1, K2a-16398A
Lengyel culture, ----, KM1, U5b1b*

N1a1a1a3, K2a-16398A distinctive Neolithic Hungary/Germany Lineages

The N1a1a1a3 and K2a-16398A samples directly link Lengyel Poland to Neolithic Hungary and the LBK culture in general.

My database holds over 1,000 mtDNA samples from ancient Europe and about 12,000 mtDNA samples from modern Europe. And amazingly, besides a few exceptions, N1a1a1a3 and K2a-16398A reside exclusively in Neolithic Germany/Hungary.

None of my modern samples belong to K2a-16398A and only three belong to N1a1a1a3(Austria, Saudi Arabia, Yeman). Interestingly Austria also provides one of my only modern examples of N1a1a1a2, which could mean there's significant LBK-like mtDNA there. Most modern European N1a1a1 belongs to different subclades than what is found in Neolithic Germany/Hungary.

K2a-16398A and N1a1a1a3 were actually quite popular in Neolithic Germany/Hungary but apparently barely exist or don't exist at all in modern Europe.

U5b1b1, a clue to the origin of Eastern European U5b1b1a and U5b1b1[c]?

The only Mesolithic sample from this study belonged to U5b1b1. The meaning of this result goes beyond U5=European HG, the subclade this Polish HG belonged gives more in-depth insight than that.

In the past year dozens of Mesolithic mitogenomes have been sequenced but U5b1b1 has popped up in only one sample: Kretuonas4, an individual from the Narva culture who belonged to U5b1b1a.

U5b1 frequencies varied in Mesolithic Europe. It looks like it peaked in the Narva culture and in Germany/France.....
     
Ukraine(39) U5b1=0
Serbia/Romania(39) U5b1=2
Germany/France(23) U5b1=7
Narva Lithuania(17) U5b1=5
Mesolithic Latvia(19) U5b1=1
Comb Ceramic(12) U5b1=1

There are at least five U5b1 clades unique to Eastern Europe today; U5b1b1a, U5b1b1[c], U5b1b2, U5b1b[3], and U5b1e1[b],

10% of Finns belong to those subclades and basically all Finnish U5b falls under those U5b1 subclades. 41% of all my U5b samples from NorthEastern Europe belong to those U5b1 clades, that's 66 out of 160 samples.

So I'm thinking that Mesolithic Eastern Europe harbored U5b1 and especially U5b1b rich WHG populations which were later absorbed by farmers and herders.