View Full Version : Origin and diffusion of J1-M267

24-03-21, 15:22
Well, finally.

See:Hovhannes Sahakyan (https://www.nature.com/articles/s41598-021-85883-2#auth-Hovhannes-Sahakyan) (Maitspalu and Doron Behar are involved)

"Human Y chromosome haplogroup J1-M267 is a common male lineage in West Asia. One high-frequency region—encompassing the Arabian Peninsula, southern Mesopotamia, and the southern Levant—resides ~ 2000 km away from the other one found in the Caucasus. The region between them, although has a lower frequency, nevertheless demonstrates high genetic diversity. Studies associate this haplogroup with the spread of farming from the Fertile Crescent to Europe, the spread of mobile pastoralism in the desert regions of the Arabian Peninsula, the history of the Jews, and the spread of Islam. Here, we study past human male demography in West Asia with 172 high-coverage whole Y chromosome sequences and 889 genotyped samples of haplogroup J1-M267. We show that this haplogroup evolved ~ 20,000 years ago somewhere in northwestern Iran, the Caucasus, the Armenian Highland, and northern Mesopotamia. The major branch—J1a1a1-P58—evolved during the early Holocene ~ 9500 years ago somewhere in the Arabian Peninsula, the Levant, and southern Mesopotamia. Haplogroup J1-M267 expanded during the Chalcolithic, the Bronze Age, and the Iron Age. Most probably, the spread of Afro-Asiatic languages, the spread of mobile pastoralism in the arid zones, or both of these events together explain the distribution of haplogroup J1-M267 we see today in the southern regions of West Asia."

Very, very interesting graphics.


I haven't yet read the paper, but off hand, the P58 clade seems very low in Europe, and extremely low in Italy. That cluster around eastern France is interesting. If it were modern North Africans etc, one would expect a similar cluster around Marseilles, if they did indeed just use hospital records or something.

24-03-21, 17:56
Some apt quotes, so we're all on the same page, as it were:

"The story of J1 or J-M267 “starts ~ 20 kya,which fits the late LGM timeframe "

" Our analyses provide better statistical confidence for the origin of haplogroup J1-M267 in the area encompassing the Caucasus, the Armenian Highland, and the Zagros Mountains."

I think most of us had already surmised this.

The frequency of modern variation and the phylogeny of theJ1a1b-Z18375 branch suggest an origin in the Caucasus or the immediatevicinity. Compared to other Caucasian lineages (J1b-F4306 and J1a2-PH77), thisbranch diversified later: in the Chalcolithic or the Bronze Age ~ 6.5 kya (95%HPD = 5.0–8.0 kya). Interestingly, an ancient individual from the Caucasus55 (https://www.nature.com/articles/s41598-021-85883-2#ref-CR55) belongingto J1a1b-Z18375 is found in the assemblage of the Bronze Age Kura-Araxescultural tradition56 (https://www.nature.com/articles/s41598-021-85883-2#ref-CR56).This cultural tradition probably originated in the Caucasus and may explain theradiation of this branch. This finding is consistent with the autosomal DNAanalysis of the Anatolian Bronze Age individuals who trace ~ 32% of theirancestry ultimately from the Caucasus hunter-gatherers/Iranian farmers57 (https://www.nature.com/articles/s41598-021-85883-2#ref-CR57).The presence of the ancient Levantine J1a1b-Z18375 individuals58 (https://www.nature.com/articles/s41598-021-85883-2#ref-CR58) andthe sub-branch composed by an Assyrian, an Iraqi, and an individual withTurkish ancestry, coalescing ~ 3.3 kya (95% HPD = 2.4–4.3 kya), may beexplained by the suggested connection of the Hurro-Urartian languages with theeastern Caucasian languages within the so-called Alarodian language family59 (https://www.nature.com/articles/s41598-021-85883-2#ref-CR59). "

As the authors point out later, they believe that J1 lineages did NOT spread into Europe during the Neolithic, but probably beginning in the Bronze Age, which would link the J1 (not P58) lineage frequency in Europe to a spread beginning perhaps in the Anatolian Bronze Age from the area either of Anatolia or the Caucasus itself. Interesting to me is the broad geographic spread in the Balkans, Spain and even France, as well as Italy.

" We conclude that haplogroup J1a1a1-P58started to diverge most probably in a region encompassing the ArabianPeninsula, the southern Levant, and southern Mesopotamia. The small-scalemigration of people from northern West Asia to these regions among whom J1a1a1-P58evolved, obviously occurred between the origins of J1a1a-Z2359 and J1a1a1-P58.The most conservative estimate is 7.4–15.7 kya."

Recent indeed.

The following is of relevance to a sample found in the Chalcolithic Levant.

"J1a1a-Z2359 is found in the Chalcolithic northern Levant close to Anatolia and among the Copper Age/Eneolithic and Early Bronze Age individuals in Central Asia, adjacent to the present-day northern border with Iran45 (https://www.nature.com/articles/s41598-021-85883-2#ref-CR45),46 (https://www.nature.com/articles/s41598-021-85883-2#ref-CR46). All these individuals share genome-wide ancestry with either Anatolian or Iranian Neolithic/Mesolithic populations. The early presence of the J1a1a-Z2359 branch adjacent to present-day northern Iran suggests the origin of this branch in the northern West Asian regions."

The following is very important.

"The expansion of haplogroup J1-M267 occurredover a long period—spanning the Chalcolithic, the Bronze Age, and the Iron Age.Many demographic events in the region could maintain the uninterruptedexpansion of this haplogroup. During the Chalcolithic and the Bronze Age,people were moving intensively across West Eurasia21 (https://www.nature.com/articles/s41598-021-85883-2#ref-CR21),22 (https://www.nature.com/articles/s41598-021-85883-2#ref-CR22),57 (https://www.nature.com/articles/s41598-021-85883-2#ref-CR57),63 (https://www.nature.com/articles/s41598-021-85883-2#ref-CR63),64 (https://www.nature.com/articles/s41598-021-85883-2#ref-CR64). At this time, the Levantine Neolithic ancestryincreased in the northern areas of West Asia22 (https://www.nature.com/articles/s41598-021-85883-2#ref-CR22), where the shared Caucasushunter-gatherer/Iranian Neolithic ancestry or the Anatolian Neolithic ancestrywere prevalent before. At the beginning of the population expansion, peoplebelonging to haplogroup J1a1a1-P58, probably migrated also to the northernregions of West Asia and Europe from the Arabian Peninsula, southernMesopotamia, and the Levant. Therefore, in the case of West Asia thisevidence—based on the TMRCAs of the shared J1a1a1-P58 branches—mirrors that ofgenome-wide ancestry22 (https://www.nature.com/articles/s41598-021-85883-2#ref-CR22). The migration of the J1a1a1-P58 lineages,though, was less pronounced towards the northern regions of West Asia andEurope, since the frequency of this haplogroup and the number of such branchesare low there. During this time and especially thereafter, the spread withinthe Arabian Peninsula, southern Mesopotamia, and the southern Levant was moreintense, resulting in a large number of local branches and the high frequencywe find today. This expansion resembles the spread of Afro-Asiatic languages inWest Asia65 (https://www.nature.com/articles/s41598-021-85883-2#ref-CR65). Both the spread of J1a1a1-P58 and Afro-Asiaticlanguages could have been caused by the change of climatic conditions and theemergence of arid pastoralism as suggested earlier23 (https://www.nature.com/articles/s41598-021-85883-2#ref-CR23)."

I had hoped that the analysis of P58 would help define the origin of the P58 samples found in Southern Europe. However, given the above statement, while perhaps it might be said that the probability might be higher than the origin of P58 samples in Europe was from the Levant, it is possible it was carried by people from, say, Anatolia, Cyprus, even the Greek islands.

"The Mediterranean and southeastern regions contain more members of haplogroup J1-M267 than the other regions. The distribution was originally associated with the spread of farming from the Near East37 (https://www.nature.com/articles/s41598-021-85883-2#ref-CR37). So far, no aDNA study has reported this haplogroup in the European Neolithic, Chalcolithic, or Bronze Age assemblages (Supplementary Table S3 (https://www.nature.com/articles/s41598-021-85883-2#MOESM7)). It appeared in Mediterranean Europe only in the historical period48 (https://www.nature.com/articles/s41598-021-85883-2#ref-CR48),66 (https://www.nature.com/articles/s41598-021-85883-2#ref-CR66). Unexpectedly, a member of haplogroup J1-M267 is found among eastern hunter-gatherers from Karelia, Northeast Europe living ~ 8.3 kya67 (https://www.nature.com/articles/s41598-021-85883-2#ref-CR67). This branch is absent in other ancient European hunter-gatherers (Supplementary Table S3 (https://www.nature.com/articles/s41598-021-85883-2#MOESM7)). Unfortunately, we fail to put this sample in the context of the current haplogroup J1-M267 variation because of the poor quality of the DNA sequence.

Although they don't seem to think so, I believe the above is a potential problem.

"A recent study with the extant variationconcludes that Asia Minor is less likely to be the source of the Greek andItalian haplogroup J1-M267 chromosomes as they do not coalesce together beforethey coalesce to the lineages from Turkey31 (https://www.nature.com/articles/s41598-021-85883-2#ref-CR31). They propose a neighboring area—theCaucasus—as the putative source. "

Well, I know they had to mention the study, but there is a problem: what did they do, fly over all the territory in between?

"The current distribution in Europe is mostlikely the result of complex demographic processes, involving various sourcesconsidering the different lineages found there. In our whole Y chromosome tree,European haplogroup J1-M267 lineages coalesce to others after ~ 5 kya.Therefore, the ancestors of the current haplogroup J1-M267 members migrated toEurope after the Neolithic, and, given the aDNA results, most likely also afterthe Bronze Age, since the earliest representative was found in the Punic periodin Sardinia at ~ 2.4 kya48 (https://www.nature.com/articles/s41598-021-85883-2#ref-CR48),68 (https://www.nature.com/articles/s41598-021-85883-2#ref-CR68) (Supplementary Fig. S2 (https://www.nature.com/articles/s41598-021-85883-2#MOESM2), Supplementary Table S3 (https://www.nature.com/articles/s41598-021-85883-2#MOESM7))."

Given how few samples we have from Bronze Age Southern Italy, Sicily, Greece, etc., I'm not sure I buy that completely.

Here, we confirm this Cohen-specific branch inhaplogroup J1-M267 as J1a1a1a1a1a1a2-B877 (Fig. 2 (https://www.nature.com/articles/s41598-021-85883-2#Fig2), Supplementary Fig. S1 (https://www.nature.com/articles/s41598-021-85883-2#MOESM1)). All Jewish lineages of haplogroup J1-M267 fallinto the J1a1a1-P58 branch (Supplementary Fig. S1 (https://www.nature.com/articles/s41598-021-85883-2#MOESM1)), which suggests their origin ultimately in theLevant. It is surprising to find two Jewish or close to Jewish J1a1a1-P58lineages in the ancient Roman samples (~ 1.5–2.0 kya)48 (https://www.nature.com/articles/s41598-021-85883-2#ref-CR48). This tells us about the migration of the Jewishpeople, at least of the bearers of the J1a1a1-P58 chromosomes, who travelledfrom the Levant to Europe via Italy, consistent with an earlier research29 (https://www.nature.com/articles/s41598-021-85883-2#ref-CR29)."

"A notable exception is J1a1a1b1-L817, a sub-branch of J1a1a1b-Z18315; while its modern members are found among the Ashkenazi Jews and central and northeastern Europeans, it is also found in a 300–500 CE individual from Rome48 (https://www.nature.com/articles/s41598-021-85883-2#ref-CR48). The TMRCA of J1a1a1b1-L817—1.5 kya (95% HPD = 0.9–2.1 kya) (Table 1 (https://www.nature.com/articles/s41598-021-85883-2#Tab1))—corresponds to the age of other Ashkenazi Jewish Y chromosome founders29 (https://www.nature.com/articles/s41598-021-85883-2#ref-CR29)."

Well, Behar has always been extremely fond of the hypothesis that Jews from Rome moved north toward the Rhine, the origin area of the Ashkenazim. :) Perhaps that is part of the fact that the "tail to the east" disappeared from Rome.


For crying out loud, given that a lot of samples are available, why would they stick with 1000 Genomes?

Somebody tell me I'm wrong.