2026 Split in the R1a-Y2619 Level

In October 2026, the Jewish Genetic Discovery Foundation reported that newly received Big Y results for a Pamiri man from Tajikistan split the R1a-Y2619 level. The Pamiri man is positive for six of the 20 SNPs (including Y2619) that had previously defined the R1a-Y2619 Ashkenazi Levite cluster, but is negative for the other 14 SNPs (including M12205) that had defined that cluster. 

As a result, the Y-DNA cluster that had been referred to as the R1a-Y2619 Ashkenazi Levite cluster for at least a decade should now more properly be referred to as the R1a-M12205 Ashkenazi Levite cluster. Indeed, FTDNA is now using the designation R-M12205 to refer to the men whose Y-DNA haplogroup FTDNA had previously reported as R-Y2619 based on Family Finder or other autosomal testing. 

On October 14, 2026, Family Tree DNA’s Discover Tool first provided an estimate of the date when R1a-M12205 split from R1a-Y2619. According to the Discover Tool's Haplogroup Report for R-M12205, R1a-M12205 began branching in 600 BCE  – 700 years after R1a-Y2619 began branching in 1300 BCE. 

Before the identification of the R1a-M12205 branch of the R1a-Y2619 line, there was an estimated 2,050-year bottleneck in the R1a-Y2619 line, starting at its formation in about 1300 BCE and continuing until the line began branching in about 750 CE. By contrast, the estimated bottleneck in the R1a-M12205 line is 1,350 years – 700 years shorter than the bottleneck previously had been.  

This new split in R1a-Y2619 provides further insight into when and where the R1a-M12205 Ashkenazi Levite progenitor may have entered the Jewish population. The 2013 Rootsi & Behar paper and the 2017 Behar paper discussed here and here provided information concerning the modern-day geographic distribution of the clusters immediately downstream from the R1a-CTS6 level – the R1a-Y2619 cluster and two brother clusters at the same level as R1a-Y2619. Those papers concluded, based on the geographic distribution of the two brother clusters to R1a-Y2619, that the R1a-Y2619 Ashkenazi Levite progenitor was most likely to have lived in the region of Iran about 3,000 years ago. The geographic distribution of those clusters reported more recently by FTDNA and YFull is generally consistent with that reported in the 2013 Rootsi & Behar paper and the 2017 Behar paper, but the existence of R1a-CTS6* men from Armenia, Azerbaijan, Iraq, and Turkey suggests the possibility that branching in R1a-CTS6 might have started, if not in Iran, somewhat to the north or west of Iran.

The new test results, for an R1a-M12205 progenitor estimated by FTDNA to have lived 700 years more recently than the R1a-Y2619 progenitor, tend to place the origins of the Ashkenazi Levite line more firmly in the area of Iran. This is because the Pamiris, also known as Badakhshanis, are of Iranian origins; their languages belong to the Eastern Iranian language family. However, the geographic location of a single R1a-Y2619 descendant is by no means dispositive of where the R1a-Y2619 progenitor lived 3,300 years ago or where the R1a-M12205 progenitor lived 2,600 years ago.  

If the R1a-M12205 progenitor lived in the area of Iran as of 600 BCE, it is unlikely that the line had entered the Jewish population as of that time. This chronology and geography would be consistent with the hypothesis, proposed by Michał Milewski, that the R1a1a Ashkenazi Levite progenitor may have entered the Jewish – and Levite population – with the Nethinim, who moved to Jerusalem with Ezra upon the return from the Babylonian Exile in the fifth century BCE as Temple servants and who were gradually assimilated into the Levite population.  

Background

Going back as far as November 2014, when this website posted its first SNP-based tree for the Y-DNA cluster referred to at that time as the R1a1a Ashkenazi Levite cluster, there had been a very long bottleneck at the R1a-Y2619 level – the level that, as of that time, characterized the R1a1a Ashkenazi Levite cluster. As of that time, the tree posted on this website identified 22 SNPs at the Y2619 level, with calculations indicating that the branch formed about 3,090 to 3,282 years before the present and began branching at about 1,346 years to 1,590 years before the present.

The 2017 Behar paper discussed here found that the R1a-Y2619 branch formed about 3,143 years before the present and began branching about 1,743 years before the present.

As of October 2026: (1) FTDNA’s Discover Tool's Haplogroup Story for R-Y2619 reported that the R1a-Y2619 branch was comprised of 20 SNPs and that it formed in about 1300 BCE and began branching in about 750 CE; and (2) YFull’s YTree for R-Y2619 reported that the R1a-Y2619 branch was comprised of 29 SNPs and that it formed about 3,000 years before the present and began branching about 1,250 years before the present. 

Before branching was identified in the R1a-Y2619 level, some insight into the likely geographic origins of the R1a-Y2619 Ashkenazi Levite progenitor could be gleaned from the modern-day distribution of the branches that are most closely related to R1a-Y2619 – the two branches that, along with R1a-Y2619, are one branch downstream from the R1a-CTS6 level. However, that distribution provides information that dates back to the time when R1a-CTS6 began branching – estimated by FTDNA to be 1300 BCE as of October 2026.   

Previous Studies Concerning the Geographic Distribution of R1a-CTS6/M582

The 2013 Rootsi & Behar Paper. The 2013 Rootsi & Behar paper discussed here identified six SNPs - including R1a-M582 (which is at the same level as R1a-CTS6) - that are characteristic of R1a1a Ashkenazi Levites and their closest matches. Rootsi & Behar identified M582 in various R1a1a populations, "with the highest frequency occurring within Iranians collected from the southeastern Kerman population who self-identified as Persians, northwestern Iranian Azeri, and in Cilician Anatolian Kurds, at 2.86%, 2.50% and 2.83%, respectively." 

Rootsi & Behar stated that "the previously proposed Eastern European origin of this lineage is no longer tenable given that our data suggest haplogroup R1a-M582 actually originated in the Near East." They noted that "[t]he higher R1a-M582 diversities and frequencies observed among Near Eastern populations indicate R1a-M582 originated in this geographic region." They concluded that "the current data are indicative of a geographic source of the Levite founder lineage in the Near East and its likely presence among pre-Diaspora Hebrews," and that "haplogroup R1a-M582 was likely carried into Europe by Jewish migrants and that it expanded among Ashkenazi Levites during their subsequent Diaspora period."

The 2017 Behar Paper. The 2017 Behar paper discussed here concluded that the R1a-M582 cluster has origins in the Middle East about 3,000 years ago. The paper identified four R1a-M582 subclusters, all of which shared a direct male ancestor who lived about 3,000 years ago: (1) the R1a-Y2619 subcluster, comprised of Ashkenazi Levites; (2) a 1,200-year-old subcluster that included an Iranian Kerman man, an Iranian Azeri man, and a Yazidi man; (3) a 2,700-year-old subcluster that included an Iranian Azeri man and an Iberian man; and (4) a line, diverging from the other lines about 3,000 years ago, that included an Iranian Kerman man. The paper concluded that the fact that each of the three clusters other than the R1a-Y2619 cluster included men with geographic origins in Iran suggests strongly that the R1a-M582 progenitor lived in that area.

FTDNA-Reported Geographic Distribution of R1a-CTS6. As of February 2026, the FTDNA Discover Tool's Haplogroup Report for R-CTS6 reported the following information concerning the geographic distribution of R1a-CTS6 (equivalent to R1a-M582) as a whole.

Because R1a-Y2619 did not begin branching until 750 CE and its geographic distribution closely matches the geographic distribution of the other major Ashkenazi Y-DNA clusters, the geographic distribution of R1a-Y2619 does not provide insight into where the R1a-Y2619 progenitor might have lived during the approximately 2,000- year period between the time when the R1a-Y2619 cluster formed and the time when it began branching.

FTDNA-Reported Geographic Distribution of Parallel Branches to Ashkenazi Branch: By contrast, the other two subclusters of R1a-CTS6 – R1a-S3418 and R1a-Y29653 – began branching soon after they formed – in 1100 BCE and 1150 BCE, according to the FTDNA Discover Tool’s estimates as of February 2026. As of that time, the FTDNA Discover Tool reported the following geographic distribution for those two subclusters of R1a-CTS6: 

Thus, as of February 2026, both R1a-S3418 and R1a-Y29653 had modern-day descendants in Iran (four tested men, two in each cluster), Armenia (three tested men - two in R1a-S3418 and one in R1a-Y29653), and Turkey (three tested men - one in R1a-S3418 and two in R1a-Y29653), and R1a-S3418 had a single tested man in each of Azerbaijan, Saudi Arabia, and Iraq. 

YFull Results for R1a-CTS6. As of October 2026, YFull's YTree for R-CTS6 reported the following information concerning the geographic distribution of R1a-CTS6, based on a combination of the test results from the 2017 Behar paper and for samples that were submitted to YFull: (1) an R1a-CTS7297 branch (the same branch reported by FTDNA as R1a-S3418), which both formed and began branching 3,000 years before the present, that was comprised of samples from Iran (two tested men), Armenia (two tested men), Azerbaijan (one tested man), Spain (two tested men), and Brazil (one tested man); (2) an R1a-Y296563 branch (the same designation used by FTDNA), which both formed and began branching 3,000 years before the present, that was comprised of samples from Iran (three tested men), Turkey (three tested men), and Iraq (two tested men); (3) a basal sample at the R1a-CTS6 level from Iran (whose results were reported in the 2017 Behar paper); and (4) the R1a-Y2619 Ashkenazi branch, which formed 3,000 years before the present and began branching 1,250 years before the present, that was comprised almost exclusively of men with reported direct male ancestry from Europe.

Interpretation of Geographic Distribution of R1a-CTS6. The geographic distribution of R1a-CTS6 has been interpreted as indicating that R1a-CTS6 likely began splitting in or near Iran and, thus, that the R1a-Y2619 progenitol likely lived in or near Iran. 

However, for more than a decade, researchers of the R1a-Y2619 Ashkenazi Levite cluster have hoped to find new branching within that cluster, recognizing that downstream branching would provide insight into the likely geographic origins of that cluster and, perhaps, when and how that cluster may have become Jewish (and Levite).

The New Test Results

In about July 2026, a member of the Jewish Genetic Discovery Foundation reviewing the Tajik DNA project page found a Pamiri man from Tajikistan who was identified as R-Y2619 based on a Family Finder test. His SNPs as reported on the project page indicated that he was positive for only one or two SNPs that were then at the R1a-Y2619 level – i.e., that he split the R1a-Y2619 level. The Jewish Genetic Discovery Foundation arranged for Big Y testing of the Pamiri man, in partnership with the Tajik DNA Project, the Afghan-Pakistan DNA Project, LeviteDNA.org, and Family Tree DNA.

In early October 2026, Big Y results were received for the Pamiri man. He was positive for six - and negative for 14 - of the 20 SNPs that had previously been at the R1a-Y2619 level. That resulted in the long-awaited split in the R1a-Y2619 level: (1) an R1a-Y2619* branch, currently comprised of the single Pamiri man; and (2) an R1a-M12205 branch, comprised of all of the men who had previously been identified as R1a-Y2619 Ashkenazi Levites. As of October 14, 2026, the FTDNA Discover Tool's Haplogroup Story for R-Y2619 estimated that R1a-Y2619 formed in 1300 BCE and began splitting in 600 BCE. 

The Pamiri man is identified on the FTDNA Tajik and Pamiri project pages as Kholinov (kit no. BP89139) from Tajikistan, with ancestral information of Tajik, Stargh, Vanj and a Y-DNA haplogroup of R-Y2619.

There are two men on the FTDNA Pamiri project page whose 12-marker results are within a step or two of the R1a-Y2619 Ashkenazi Levite mode at 12 markers – close enough to make it appear likely (but by no means certain) that they belong to the same R1a-Y2619* cluster as Kholinov; those two men are identified on the Pamiri project page as R-M512, based on Y-DNA12 testing. Those two men, both identified as being from Tajikistan, are identified as Nematuloev (kit no. 343354) and Shakurov (kit no. 343356), both with ancestral information of Yazgulami, Zhamag. (Big Y testing on those men would likely provide downstream branching within the Pamiri branch of R1a-Y2619; if so, such testing would identify a terminal SNP for the Pamiri branch and allow calculation of the age of that branch.)

Stargh and Zhamag are both villages in the Vanj District of the Gorno-Badakhshan Autonomous Region, located in eastern Tajikistan just to the north of Afghanistan.  

Stargh is located at 38.5982N 71.8246E at the altitude of 6,968 feet (2,123 meters), and Zhamag is located at 38.2728N 71.5705E at an altitude of 6,709 feet (2,044) meters. Stargh is an entry point to the Yazgulom Valley, and Zhamag is located within the Yazgulom Valley. On the Pamir Trail, the hike from Stargh to Zhamag is three stages of 43.8 kilometers, 10.2 kilometers, and 10.3 kilometers, for a total of 64.3 kilometers; the second phase has an ascent of 2,163 meters and a descent of 40 meters, while the third phase has an ascent of 524 meters and a descent of 2,401 meters. Approximately 741 people live within a seven-kilometer radius of Stargh, and approximately 745 people live within a seven-kilometer radius of Zhamag.

The Yazghulami people are an ethnic minority group native to the Yazgulom River valley. Their native language, also named Yazghulami, is a member of the southeastern subgroup of the Iranian language and was spoken by about 4,000 people as of 1994.

As of 2022, there were about 230,000 Pamiris in Tajikistan. There are seven dialects of Pamiri. Only about 3 percent of Pamiris speak the Yazghulami dialect of Pamiri.

In the mid-6th century BCE, Cyrus the Great of the Achaemenid dynasty conquered much of what is now Tajikistan. In 330 BCE, Alexander the Great conquered the Achaemenid Empire, who annexed it to his Macedonian Empire. The direct male ancestor of the Pamiri R1a-Y2619 line may have arrived in Tajikistan from elsewhere in the Achaemenid Empire during the two centuries of Persian rule over Tajikistan. 

Such a timeframe would match nicely with a split between the Pamiri R1a-Y2619* branch and the R1a-M12205 Ashkenazi Levite branch. It would also tend to support the likelihood that the progenitor of the R1a-M12205 Ashkenazi Levite branch lived in Iran as of 600 BCE, which would in turn tend to suggest that the R1a-M12205 line was not part of the Jewish population as of that date.

This timing also tends to support a hypothesis, previously advanced by Michał Milewski, that the progenitor of what is now known as the R1a-M12205 line might have been a Nethinim who returned to Jerusalem with Ezra at the end of the Babylonian Exile.  

However, the geographic location of a single R1a-Y2619 man in the present day is by no means dispositive as to where the R1a-M12205 progenitor lived 2,600 years ago.

Current Branching Within R1a-Y2619 

The following six SNPs are now at the R1a-Y2619 level: (1) Y2619; (2) FGC10618; (3) FGC10619; (4) FGC10625; (5) FGC10629; and (6) M12210.

The following 14 SNPs are now at the R-M12205 level, one level downstream from the R-Y2619 level: (1) M12205; (2) FGC10617; (3) FGC10620; (4) FGC10621; (5) FGC10624; (6) FGC10626; (7) FGC10633; (8) FGC22265; (9) M12200; (10) M12204; (11) M12211; (12) Y2739; (13) Y2741; and (14) YP466.

As more Big Y results are received (and, perhaps, as Y-DNA tests are performed on more ancient burials), it is possible that additional branching will be identified within the R-M12205 level. Any such results would be likely to provide further insight into when and how the R-M12205 line became Jewish (and Levite).