Y HAPLOGROUP L

Y-haplogroup L is common (at frequencies around 15%) in India (in particular the South) and in Pakistan. It appears at low frequencies (2-4%) also in the Northern part of the Middle East (Iran, Turkey, Armenia, Kurdistan, Lebanon, but much less among Arabs) and perhaps in Central Asia (Tajikistan and Uzbekistan, though these regions have not been well tested). A couple of papers, as well as a number of personal tests, have shown the presence of haplogroup L in the Mediterranean, although at very low frequencies (1% or less). It is extremely rare in Northern Europe and America, but it does pop up now and then north of the Mediterranean (e.g. Switzerland and Germany).

The latest nomenclature and SNPs related to the group can be found on Isogg, as well as on the Y-Haplogroup L Project at FTDNA. Most members of the L haplogroup belong to the first group, L1, itself divided into L1a (mostly India and Pakistan), and L1b (Middle Eastern, Central Asia, Europe). The few people who do not belong to L1 most likely belong to another group, L2. The discovery of the structure of the L tree is relatively recent. The change in nomenclature can generate a lot of confusion and make it hard to identify the correct SNP, especially when reading older papers.

Sengupta et al. (see below) was perhaps the first study that tried to characterize the structure of the haplogroup. It isolated the three largest subgroups of haplogroup L, M76-M27, M317 and M357, and named them L1, L2, and L3. In the latest tree, these correspond to L1a1, L1b, and L1a2. Later, personal tests identified a few people who did not belong to any of these subgroups; some of them were found to belong to a different subgroup, characterized by a new SNP L595, now defining the L2 subgroup.

M76 (current L1a1, former L1) is the most common subgroup in India, while M76 and M357 (current L1a2, former L3) have approximately equal weight in Pakistan. M317 (current L1b, former L2) is rare in the Indian subcontinent. Iran seems to have all three major subgroups, while Turkey appears primarily M357. Other papers have found additional markers. For instance, L1b can be divided into two subgroups, M247 and M349. The people who do not belong to L1 have not been studied in academic papers, but only in personal genetic tests. Their ancestry is European, but it is possible that this group is present in the Middle East or Caucasus, where few people have tested.

Unfortunately, few papers (except that of Sengupta on India) have examined the origin and spread of the haplogroup.

My result is a little surprising because my family (traced back at least to the XV century) is from the area of of Massa, in Northern Tuscany, and the majority of Northern and Central Italians belong to haplogroup R1b (and if not to E3b and J2). But the town is on the coast, and was occupied by many invaders over time, including Byzantines (that is, from the Eastern Mediterranean).

Some links about Y haplogroups in general:

• Y-haplogroup L project, managed by Gareth Henson and Peter Hrechdakian. I strongly recommend L people to join the group, which is providing useful insights about the distribution of the haplogroup.
• Ethnoancestry. The company specializes in SNP tests and offers, among others a multiplex test to find out the haplogroup directly, that is, without going through STRs. (I did it. My test was actually done by Marligen). EA also offers many subclade tests, including one for haplogorup L.
• The Genographic Project, which I used too. They start with an STR test, which is not very informative for haplogroup L at this point, with so few observations. So far, however, the Genographic Project offers the cheapest test. Also note that, while their general description of haplogroups is easy and informative, it has not been updated. Regarding haplogroup L, they overestimate its frequency in Southern India, and do not talk about its more likely origin in the Pakistan/NW area, where it has more diversity.
• FTDNA, which specializes in STR tests (and in fact does the work of the Genographic Project).
• ysearch, a database of Y-chromosome self-reported STR and SNP's, created by FTDNA. I uploaded my results there too. One can compare one's own sequence with that of the other people in the list. Unfortunately, there aren't very many L people, but some are slowly appearing.
• Yhrd, a large collection of Y-STR haplotypes, where one can check if his sequence has matches (mine hasn't).
• YCC NRY tree 2002, the tree diagram of the various haplogroups. (as of 2002. New branches may have been found). Useful because it lists the various names that have been given to the different haplogroup in early papers, which did not follow a unified nomenclature yet.
• The tree is explained in the review article The human Y chromosome: an evolutionary marker comes of age by Jobling and Tyler-Smith, and more recent findings on various branches are collected in the tree posted by the International Society of Genetic Genealogy (ISOGG).
• Whit Athey's haplogroup predictor, which predicts haplogroup from STR data. The new version includes haplogroup L.

• Polarity and temporality of high resolution Y-chromosome Distributions in India Identify Both Indigenous and Exogenous Expansions and Reveal Minor Genetic Influence of Central Asian Pastoralists, by Sengupta et al. (American Journal of Human Genetics 78(2): 202–221) studies a large sample of Indians and talks a little about haplogroup L. The paper is difficult to read, but has some very interesting information. First, it shows the frequency of haplogroup L in India and Pakistan, respectively 6.7% and 13%. This is lower than what shown in other papers, in part because the sample per se is not representative, tribes (which are a minority of the population) are oversampled. More interestingly, the article isolates three subgroups of haplogroup L. They are characterized as follows: L1 M76, L2 M317, L3 M357 (as explained, the names are now different in ISOGG.). The latter two are novel markers. Interestingly, the Indian sample is predominantly L1, while the Pakistani L's are more or less half L1 and half L3 (plus a couple L2 among the Makrani), which suggest that the haplogroup may have originated in the NW/Pakistan area rather than in Southern India. Unlike some previous papers, the group seems overall more frequent, and more dispersed, in Pakistan than in India. The paper also shows the frequency in various Indian groups. Dravidian caste around 17%, Indo-European caste 2-4%, Dravidian tribe 5%.
• Ethnoancestry has just introduced a test for subgroups of L, which tests M76, M317 and M274, M357 and PK3, M349. Some people, including myself, have already received their results, which are generating interesting information.
• The international society of genetic genealogy (ISOGG) has posted a very recent Y-dna tree, including a branch for haplogroup L. In this tree, haplogorup L is defined by a series of SNPs (M11 and M20 the usual tested ones). L1, L2 and L3 are as in Sengupta. PK3 is considered a subgroup of L3 and called L3a.
• Excavating Y-chromosome haplotype strata in Anatolia by Cinnioglu et al. Interesting because it is the best paper on haplogroup L outside of the subcontinent. His sample contains 22 L people, whose STR sequences are also shown. There isn't much discussion about the L group, only one paragraph about the fact that the Turkish sample seems to miss the M27 (ie Sengupta's L1) mutation present in India and Pakistan. He also has one M349 observation, which he calls L2.
• Turktoresi has a list of papers reporting frequencies of haplogroups, including L, in Turkey. In Turkish, but the frequencies are easily understandable.
• Y-chromosomal diversity in Lebanon is structured by recent historical events, by Zalloua et al has a large sample from Lebanon, which allows one to estimate the frequency of L in Lebanon at around 3-5%. A supplement provides STRs. While the paper focuses on other topics, the sample is potentially important because Lebanon is a major candidate for the spread of L in the rest of the Mediterranean.

India and Pakistan
• An excel file with STRs from the papers by Qamar and by Sengupta. Useful for a general idea of different types of Ls and for STRs from the subcontinent. (Certain STRs may be incompatible between the two papers, though).
• Y-Chromosomal DNA Variation in Pakistan by Qamar et al. shows frequencies in a Pakistani sample (it's called haplogroup 28): overall 14%, and more or less everywhere throughout Pakistan. The paper mentions the possibility that the L group is related with the arrival of agriculture in the area.
• Mohyuddin et al "Detection of novel Y SNPs provides further insights into Y chromosomal variation in Pakistan" claims to have found a branch of L among a sample of Kalash, which they call L4, with mutation PK3. This was later shown to be subhaplogroup L3a.
• Sahoo et al A prehistory of Indian Y chromosomes, (PNAS jan 24 2006 v103(4), p.843-848) examines a larg sample of Indian Y's. The paper does not have frequencies and is quite dense. There is one map showing the frequency of haplogroup L in India and Pakistan (although its source is not clear).
• The Genetic Heritage of the Earliest Settlers Persists Both in Indian Tribal and Caste Populations by Kisivild et al. has a nice table the frequencies of the L group in his Indian samples (12% in Punjab and 29% in Tamil Nadu, less in the tribes).
• Again on India Basu et al Ethnic India: A genomic view (2003) has a table (table 2) with frequencies of L. Slightly smaller than above, around 10% Dravidian caste, 2.4 IE caste. The article doesn't focus on hap L though.
• Independent Origins of Indian Caste and Tribal Paternal Lineages by Cordaux et al. shows the frequency of group L in his sample (19% in caste, 7% in tribes).
• Y-chromosomal evidence for a limited Greek contribution to the Pathan population of Pakistan, European Journal of Human Genetics 2006, by Firasat et al. uses more or less the same data as Qamar, but shows some subtyping. The 12 Pathan L's, for instance, are 7 L3 and 5 L1 (the ysearch Pathans seem also to be L3). There is also a sample of 77 Greeks, none of whom is L.
• The genetics of language and farming spread in India by Kivisild and others. In some tables, he shows the frequency of L in his samples from India and neighboring countries.
• Thanseem et al. Genetic affinities among the lower castes and tribal groups of India: inference from Y chromosome and mitochondrial DNA, BMC Genetics 2006 examine a sample of three tribes. As shown in other papers, haplogroup L has a frequency of around 8% among the tribes, which is smaller than the frequency in the castes.
• Watkins et al. Genetic variation in South Indian caste: evidence from Y chromosome, mitochondrial and autosomal polymorphisms (BMC Genetics 2008) finds about 10% L in Tamil upper chastes, 20% in Middle castes, and 3% in lower castes (each group with about 40 observations). Mostly L1.

Middle East and Central Asia
• Regueiro et al: Iran: tricontinental nexus for Y-chromosome driven migration finds 8 (out of 140) L people in Iran, 6 of them in (unclearly defined) Southern Iran. 4 are L1, 3 L2 and 1 (the Northern guy) L3. Without any further comments, his table shows M349 and M274 as downstream markers of M317, and defining the haplogroups that he calls, respectively, L2b and L2a (the reverse of what is now used by FTDNA). Nothing else is said about the haplogroup.
• The Eurasian Heartland by Wells et al. has a nice table with frequencies in samples from central Asia (it's the M20 column). It seems the frequencies in Uzbekistan and Tajikistan are around 10%, which seem high given subsequent research. The numbers from India (up to 50%) also seem high. No discussion is made.
• Zerjal et al. A genetic landscape reshaped by recent events: Y-chromosomal insights into central Asia. (Am J Hum Genet 2002 sep 71(3)) finds almost no L in central Asia.
• Karafet et al. Paternal Population History of East Asia (Am J HG 2001) finds 3 (out of 67) among Uyghurs and 2/54 among the Uzbeks (hap L is her group 27). This suggests that there are some L in Central Asia, but may be in the order of maximum 5%. (Incidentally, Sengupta has tested one of the Uyghur observations, and it turns out to be L2).
• Armenian Y chromosome haplotypes by Weale et al. shows some samples from Armenia (frequencies around 2-4%). Again no discussion.
• Isolates in a corridor of migrations: a high-resolution analysis of Y chromosome variation in Jordan, by Flores et al. J Human Genetics 50 p 435. It has a table in which he lists frequencies for the L haplogroup from various studies. They are: Turkey 4.2%, Iraq 1, Iran 3.1 Pakistan 13.5, Lebanon 1.9, Kurdistan 3.2, Greece 1.1
• Reconstruction of patrilineages and matrilineages of the Samaritans by Shen et al. (thanks to Yuri Khripin of Marligen) shows 35% L haplogroup among a sample of Druze, but no L among Jews or Palestinians.
• The Y chromosome of Jews by Nebel et al. shows some L in Kurds but not in Jews or Arabs.
• Y chromosome sequence variation and the history of human populations by Underhill et al reports 21 observations in his worldwide sample (out of around 1000), in the Middle East, Central Asia, and above all Pakistan and India (10/88) and the Hunza Valley of Pakistan (where the language isolate Burushaski is spoken).
• Human Y chromosome variation in the western Mediterranean, by Scozzari et al. (Human Immunology 62 p871)) seems to test for it (if I understand correctly, it's called 26.2). He does not seem to find it almost anywhere in a large sample of European people (just 2?) except an interesting 25% of a 35 member Druze population. As usual, nothing at all is said about the group. Interesting the data on the Druze. The presence of L in Lebanon (outside of the Druze), is still an open question, although examples exist on ysearch.
• Zalloua et al (2008) Identifying Genetic Traces of Historical Expansions: Phoenician Footprints in the Mediterranean has some additional observations from the Levant area. As in previous papers, L is detected in Lebanon, Syria and Turkey. There are also a couple of Palestinian (the first detected, as far as I know). None however is detected in S Italy, Tunisia and W Mediterranean, and just a couple in Greece. Being present in Lebanon, L is of course a candidate for a phoenician expansion, but none was detected here in the colonies (though as indicated above some papers do find some L in Italy).
• El Sibai et al, Geographical Structure of the Y-chromosomal Genetic Landscape of the Levant: A coastal-inland contrast, Annals of Human Genetics 2009 (or 2010?) finds many L in Lebanon-Syria, especially in Eastern Syria (30%). As usual, nothing special is said about the haplogroup.
• Abu Amero et al. Saudi Arabian Y-Chromosome diversity and its relationship with nearby regions (2009) detect between 1 and 3% of L1 in Saudi Arabia, UAE, and Qatar.
• Afghanistan from a Y-chromosome perspective, European Journal of Human Genetics 2012, has some supplemental material with a table with the percentage of L in various countries (table 3), and the list of STRs for the Afghan sample (including a L*).

Europe
• The genetic legacy of paleolithic homo sapiens in extant Europeans: a Y chromosome perspective by Semino et al. Unfortunately, haplogroup L does not seem to be relevant for understanding Europe, so no discussion is made. As far as I have seen, though, it is the only paper that shows cases of L haplogroups (called here Eu17) in Italy, 2 observations (out of a sample of 37) in Calabria. Plus one in Greece (out of 37), Turkey, Lebanon, in Hungary and in Andalusia (all groups having similar sample sizes). A later paper by Di Giacomo et al examines a much larger sample from Italy, but very regrettably fails to test for the L group.
• Capelli et al. - Y chromosome variation in the Italian peninsula is clinal and supports an admixture model for the Mesolithic-Neolithic encounter - (Molecular Phylogenetics and Evolution 2007) has data on Italy (mostly central). As usual, they do not test for L but for K. However, a quick comparison with STR from other L observations suggests that 10 observations (out of 700, ie slightly more than 1%) are L. (Including, curiously 3 from Alta Val Badia, showing that L did move around a little bit.)
• Pichler et al. - Genetic structure in contemporary South Tyrolean isolated populations - also finds 3 L out of 194 people in S Tyrol (two in Val Badia), though curiously the STRs are different than those in the Capelli paper.
• A web page by Dienekes Pontikos has a summary of the distribution of Y haplogroups in Greece. Haplogroup L seems even less frequent than the 1% or so.
• Gresham et al Origins and Divergence of the Roma. This paper did not test for the L group, and L people would fall in their group VI-71 (ie M89, which defines F-R, but not a number of other loci defining other subgroups), of which they find only 3.6%. The sample covers a diverse set of Roma mostly from Bulgaria (but also from Spain and Lithuania). If the sample is representative of the entire Roma population, this suggests that few Roma are L (most are H). And interestingly, the few people falling into this VI-71 group are Turkish speaking Roma (which suggests an admixture in Anatolia, where these groups are more frequent). Similarly, a paper by Kalaydjieva et al. Pattern of inter and intra group genetic diversity in the Vlax Roma seems to find no L in her sample (although it is not clear if and how she tested).
• The genetic male legacy from El Salvador (Lovo-Gomez et al, Forensic Science International 2006) shows 2 L's in El Salvador (in a sample of 150). If true, this shows the presence, at very low frequencies, of haplogroup L among Latin Americans, something that has been documented also by some ysearch results.

I'd like to know more about both haplogroups. Do you belong to haplogroup L as well? Do you have any information about it? I'd be happy to share. Send me an email to:

cacio 'at' cagetti dot com
(type in the address with the @ in your mail program)

Back to Cacio's genetics homepage.