Last Update: Dec. 18, 2004
A recent study by Semino et al is one of the best sources of information about haplogroup J. Of special interest are the maps that show the geographical distribution of several J subgroups. In general, the J's are more common in Europe as one looks farther south and east, with the maximum concentrations near the Mediterranean coast. Some examples from Semino's study: Greece 22% (of the total population), Italy 20-30%, Spain 3-9%, the Ukraine 7.3%, Dutch 0%.
According to Al-Zahery et al both J1 and J2 expanded out from the Fertile Crescent about 7000-9000 years ago. The J2's seem to have started in the northern part of the Fertile Crescent and expanded west towards Europe (perhaps by way of the Balkans). As a result, almost all European J's belong to J2. The presence of the J2 subgroups in India, Pakistan, and Nepal means that they must also have expanded towards the east. However, at present, little is understood about the timing or dynamics of that eastward J2 dispersal.
Al-Zahery et al thought the J1's originated in the southern part of the Fertile Crescent and spread in a migration that happened after (and may have been triggered by) the earlier migrations from the north. However the J1 migrations were in more southern directions - towards the Middle East, the Arabian Peninsula, and east and north Africa.
The study by Semino et al, and another study by Cinnioglu et al, reported STR haplotypes for several J subgroups. Combining the results from those studies shows that the modal haplotype for haplogroup J is probably something like:
DYS19=14The data from these studies show that there is a lot of variation among the J haplotypes. There is enough variation that there are no hard and fast rules for using STRs to estimate haplogroup J. For example, a rough estimation of haplogroup J could be made for a haplotype with large alleles at DYS388 (15-18) along with DYS393=12. The problem is that many haplogroup J members have DYS388=14 - and many haplogroup I members have DYS388=14 or 15. Alleles of DYS393=12 are also seen in other haplogroups than J. Since many different alleles are seen at several markers in all sub-groups, it will be difficult to use STRs to estimate J sub-groups. I think it will require SNP testing.
Among all the common European Y haplogroups, J stands out as the haplogroup in which SNP testing could be most useful. Just about every J would learn something from SNP testing about the deep ancestry of his paternal line. Knowledge of Y haplogroup J has now advanced to the point that geneticists are piecing together interesting stories about some of the J subgroups. But I don't think we are going to be able to use STR haplotypes to guess at J sub-groups. Most of the subgroups' STR haplotypes are just too similar to each other. It'll require SNP testing.
A study by Paracchini et al found the frequencies of the J sub-groups seen among haplogroup J Caucasians in California and Hawaii were:
Paracchini's study looked for all the haplogroup J subgroups shown in the 2002 YCC nomenclature - but these four subgroups were the only ones they found among 200+ Caucasian men. These four J subgroups, plus J1* (M267), are probably the main J subgroups for which a haplogroup J man should be SNP tested. Most of the other J sub-groups shown in the YCC tree are too rare to be worth testing by the typical genetic-genealogy customer.
A comparison of the allele frequency graphs for J1* and J2* can be seen here.
A comparison of the allele frequency graphs for J2e1, J2f*, and J2f1 can be seen here.
More information about the J sub-groups can be found below.
About half of the J2's don't belong to any of the known J2 sub-groups - so they are classified as J2*. Eventually new markers will be found that will divide all these J2*'s into their own J2 subgroups - but in the mean time they are lumped together into the "paragroup" J2*.
The markers that define J2 subgroups are mutations that have occurred more recently in time than M172. Some of these J2 subgroups are defined by mutations that are "recent" enough that geneticists are able to decipher stories specific to those J2 sub-groups. But you'll have to wait until a marker is found for your J2 line to learn if there are any stories specific to your direct paternal line.
Until then, the best guess would be that your direct paternal ancestors were once part of the expansion out of the Fertile Crescent about 8000 years ago that followed the invention of agriculture. Since the farmers had more a more reliable source of food than did their hunter-gather neighbors, they had more children. More children meant they needed more land - so they kept growing into their neighbors' territory.
According to a 2004 study by Di Giacomo et al, the J2* paragroup can be separated into two lineages based on the alleles seen at DYS413 (an STR marker not presently tested by any commercial company). Most European J2* men will have both DYS413a and DYS413b of 18 or less. This group is believed to be more closely related to the sub-groups J2a, J2f and J2f1 than it is to the other J2's. The rest of the J2* men will have alleles at DYS413a and DYS413b of greater than 18. Di Giacomo et al also suggested that a large part of the Neolithic migration may have happened by sea rather than by overland routes.
The geographical distribution of J2* is shown in the map in Semino's paper labeled "J-M172".
The footprint of J2e1 (M102) on the European map indicates some sort of connection between the southern Balkans and north-central Italy. One possible explanation is that J2e1 may have dispersed into Europe from the Balkans, but Semino et al had no suggestions of any cultures seen in the archeological record known to connect those two regions. ( One of the E3b sub-groups also seems to have entered Europe through the Balkans.) The map by Semino et al titled "J-M102" suggests that J2e1 may have subsequently spread from north-central Italy west across Europe. Some of the highest frequencies that Semino et al saw for J2e1 were Albania (14.3% of total population), north-central Italy (9.6%), Greece (6.5%), the southern Caucasus (6.3%), Croatia (6.2%), Bernais (3.8%), and India, Nepal, Pakistan (3-8%). Raito et al found J2e's at a rate of 14% among the Saami in Kola, Russia. Semino et al estimated the date of the M102 mutation at about 8000 years ago.
J2e1 is the one J sub-group that has a distinctive enough STR haplotype that it may be possible to recognize it from the others. It differs from the usual J haplotype by DYS19=15, DYS389i=12, DYS390=24, and YCAII=19,20. Cinnioglu et al also report that J2e1 can be distinguished by its alleles at DYS461. There are some nomenclature difficulties, but I think that Relative Genetics, and DNA Heritage would (since June 2004) report that J2e1 has DYS461=9 or 10 (these values are larger by 2 than the values given in Cinnioglu's paper). Those labs would report most R1b's and I's as DYS461=12, and most other J's would have DYS461=12 or 13. Examples of haplotypes that may be J2e1 can be seen here.
Cinnioglu et al also found that more than half of the J2e1's in Turkey belonged to a J2e1 subgroup defined by M241 (not in the 2002 YCC nomenclature). Semino et al also reported what may be a sub-group of J2e1-M241 defined by M280. Their data are sparse, but M280 may be worth testing since it was only seen in Greece (2.2%).
It may also be possible to further divide the J2e1 haplogroup into two sub-lineages based on a different kind of test called 49a,f Taq/I. It's not a SNP test, but it's a different kind of Y "marker" that mutates almost as slowly as SNPs. Most J2's give a 49a,f Taq/I result called ht7. But about 2/3's of the J2e1's give a result of ht24, and about a third of the J2e1's give a result of ht8 (ht8 is also seen in most J1* M267).
Cinnioglu et al saw a distribution in Turkey of the sub-groups, J2f* (M67) and J2f1 (M92), that matched well with the territory of a complex society that flourished in NW Anatolia about 4500 years ago. The Maritime Troia Culture "engaged in widespread Aegean trade... involving both the western Anatolian mainland and several of the large islands in the eastern Aegean, Chios, Lemnos and Lesbos". The three earliest layers found in the excavations at Troy are associated with the Maritime Troia Culture. Semino et al agreed that J2f* and J2f1 could have arrived into Europe by way of "sea-faring Neolithic populations who reached southern Italy", or perhaps migrations from Anatolia via the Bosporus Isthmus. Semino et al thought that J2f* and J2f1 could represent, at least partly, the Y component that King and Underhill tied to ceramics in their recent paper. Semino et al estimated the ages of the J2f* and J2f1 subgroups at about 12,000 years and about 9000 years, respectively. These sub-groups (along with J2a and most of the J2*'s) can be distinguished by their alleles at an STR marker called DYS413 (not presently tested by any of the genetic testing companies). For these groups, DYS413a and DYS413b both have alleles of "18" or less (other J's have alleles at DYS413 of greater than "18"). Members of J2f1 may have DYS390=22 more commonly than other members of J (though DYS390=22 is also seen in haplogroups other than J).
Semino saw the greatest frequencies of J2f* in Georgia (13.3%), northern Italy (9.6%), southern Causasus(6.3%), Iraq (4.5%), Greece (4.3%), Spain (3.4%), Turkey(3%), and Albania (3.6%). The distribution of J2f1 seemed to show connections between Anatolia and southern Italy. They saw the sub-group J2f1 in southern Caucasus (6.3%), Greece(3.3%), Turkey (4.4%), Italy (3-7%), and Pakistan/India (4.5%).
The highest frequencies of J1 (M267) are seen in the Middle East, north Africa, and Ethiopia. The marker M267 was not shown in 2002 YCC nomenclature, but Semino et al use it to define J1. (M62 now defines a sub-group of J1). While not seen in Paracchini's study, J1 may be another J sub-group worth testing. Semino et al show that J1's commonly account for 10% of the J's seen in Europe. The authors claim that the J1's from the Middle East and north Africa seem to be distinguishable from the rest of J's by having YCAII=22,22. Cinnioglu's data shows that many J1's have DYS461=11 (using Relative Genetics new nomenclature). The result of a 49a,f/TaqI test for J1 is usually ht8.
Behar et al found that about 20% of Ashkenazi Jews were J1 (and another 20% were J2).
The only subgroup of J1 that might be worth testing is defined by M365 (not in 2002 YCC nomenclature). Semino saw it at rates of 1-2% in Turkey and Georgia.
If you are a J, but not a J2, then you are almost certainly a member of J1. Di Giacomo et al found only six J* men out the 376 haplogroup J men in the study.
A comparison of the allele frequency graphs for J1* and J2* can be seen here. J1* seems to mostly have DYS461=11, while J2* has DYS461=12,13.