Pages

February 9, 2014

Italian haploid genetics (second round)

More than a year ago I commented (as much as I could) on the study of Italian haploid genetics by Francesca Brisighelli et al. Sadly the study was published with several major errors in the figures, making it impossible to get anything straight. 

I know directly from the lead author that the team has been trying since then to get the paper corrected but this correction was once and again delayed by apparent inefficiency of PLoS ONE's management, much to their frustration. Finally this week the correction has been published and the figures corrected.

So let's give this study another chance:

Francesca Brisighelli et al., Uniparental Markers of Contemporary Italian Population Reveals Details on Its Pre-Roman Heritage. PLoS ONE 2012 (formally corrected in February 2014). Open accessLINK [doi:10.1371/journal.pone.0050794]
 
Notice please that you have to read the formal correction in order to access the new figures, the wrong ones are still in the paper as such. 

The corrected figures are central to the study:

Figure 1 (corrected). Map showing the location of the samples analyzed in the present study and those collected from the literature (see Table 1).
Pie charts on the left display the distribution of mtDNA haplogroup frequencies, and those on the right the Y-chromosome haplogroup frequencies.

So now we know that the Northern mtDNA pie was duplicated in the original graph and that Central Italians are outstanding in R0(xH,V), which reaches 14% (probably most HV*), while they have some other peculiarities relative to their neighbors from North and South: some less U and no detected V. 

Other variations are more clinal: H decreases from North to South while J and T do the opposite.

Figure 3 (corrected). Phylogeny of Y-chromosome SNPs and haplogroup frequencies in different Italian populations.

In the Y-DNA side, the most obvious transition is between the high frequencies of R1b1a2-M269 (R1b3 in the paper) in the North versus much lower frequencies in the South. But also:
  • J2 is notorious in the Central region (and also the South) but rare in the North.
  • G frequencies in the South are double than those of Center and North.
  • The same happens with lesser intensity regarding E1b1b1-M35 (E3b in the study).
  • In contrast haplogroup I is most common in the North. However the Sardinian and sub-Pyrenean clade I2a1a-M26 (I1b2 in the paper), which is also the one documented in Chalcolithic Languedoc, is rare in all regions.

The study also deals with several isolated populations:

Figure 4. Haplogroup frequencies of Ladins, Grecani Salentini and Lucera compared to the rest of the Italian populations analyzed in the present study.

All them show large frequencies of mtDNA H relative to their regions. The Grecani Salentini do have some extra Y-DNA E1b1b1 (E3b) and J2, what may indeed underline their partial Greek origins. The Ladini show unusually high frequencies of R1b*(xR1b1a2) and K*(xR1a,R1b,L,T,N3), while the Lucerans are outstanding in their percentage of G.

I want to end this entry with a much needed scolding to the staff of PLoS ONE for their totally unacceptable original sloppiness and delay in the correction. And my personal thanks and appreciation to Francesca Brisighelli for her indefatigable persistence and enthusiasm for her work, which is no doubt of great interest.

22 comments:

  1. The predominance of R and I in the North and Central regions make sense - other groups (E and J and partially G) apparently have affected the Mediterranean from the Neolithic to the Bronze and Iron ages.

    The G distribution again reminds us that it may have been high in the Balkans before/ just at the beginning of the Neolithic.

    It is also quite educational that mtDNA H is inversely correlated to SE European/ West Asian y-DNA groups (E, G, J - as long as R is not originally considered such, given its likely NW subcontinent/ Gravettian origin).

    ReplyDelete
    Replies
    1. I must agree with what you say this time, Eurologist. I'm also pondering the issue of Y-DNA G (G2a-P303 for all I know), whose main distribution is Southern Italy, Sardinia, Iberia and areas North and West of the Alps. Further East it seems rare, excepted some NW Caucasian populations.

      I would think that the main center of expansion was Southern Italy in the context of Cardium Pottery and maybe it had a less significant impact in the Danubian Neolithic, which in any case would seem to require another origin (unless it arrived there via La Hoguette culture). It is strange indeed that, lacking a clear Balcanic origin (as happens with the other Neolithic lineages), G2a would end up so homogeneous in Europe West of the Adriatic.

      Delete
  2. This clearly proves the messapic with G was epirote or greek and not illyrian as stated by modern historians

    ReplyDelete
    Replies
    1. It rather would seem to prove it Italian, right?

      There's more G in South Italy than in all the Balcans and for what I can discern, the haplogroups are different: Balcanic G tends to be akin to Anatolian G, while Italian G is generally of the same subclade P303 as found in other parts of Western Europe, notably Iberia and across the Alps.

      Almost certainly Southern Italy is the area with more yDNA G of all Europe, of the Western variant P303 or mostly so, which is seldom found in the Balcans.

      Delete
  3. This clearly proves the messapic with G was epirote or greek and not illyrian as stated by modern historians

    ReplyDelete
  4. The interesting points for me are the Messapics have G, matching greek and epirote people, as well as J2 etc.
    and
    K2 (T ) and L is split , yet they still have K as well............must be very old

    ReplyDelete
    Replies
    1. You seem the same person as Victor Pretotto, right?

      That K* is indeed interesting because nearly all K in West Eurasia is P or LT and these have been excluded. It's found in several populations across Italy: Ladini, Latini and Pugliani. I now wonder if the excess K(xP) found in Eivissa (Ibiza) and some other scattered Iberian populations could also be this kind of K* and not necessarily all T. I say, considering the obvious Ibero-Italian relation in other Neolithic lineages, notably G2a.

      Now that this issue arose, I notice that the "Sanniti" (Samnites, i.e. modern populations of inland Southern Italy, near Naples) have high frequencies (10%) of I-M26, as do Sicilians (7%) being the only populations of Italy with greater frequency of this I subclade than others. Again this strongly suggests that Southern Italy was a key hotspot in the spread of Impressed-Cardium Pottery westwards. Nothing new in archaeological terms but it's interesting to see it so clearly reflected in genetics: both I-M26 and G2a (with some E-V13 as well) were the main Neolithic lineages that reached further West into Sardinia-Corsica and, Occitania and Iberia. Compare with Adams 2008: http://www.cell.com/AJHG/fulltext/S0002-9297%2808%2900592-2

      Delete
    2. The sanniti would incorporate also J2 Corinthian Greeks , ie from Ancona. The again the only Greeks in Italy where Corinthian Greeks.
      Picene would also have a gallic influence as noted by Roman historians. The Picene breakup is indeed interesting, looking more western than eastern

      Delete
    3. J2 in Greece is more of a Cretan (and Larissa) Greeks. Other Greeks do not have notable amounts of J2, except Epirotes (different subclade partly) and some Asian islanders (Chios). See HERE.

      Also I see no particular reason why the Samnites would be influenced by Greece.

      Finally I must dispute your claim that "the only Greeks in Italy where Corinthian Greeks". That's simply not true, see: http://en.wikipedia.org/wiki/Magna_Graecia

      The only Corinthian colonies I could find are Corcyra (Corfu) and Naucratis (in Egypt). It does not seem that Corinth was a particularly active colonial power, rather the colonies of Magna Graecia were largely founded by Achaeans (Patras) and Ionians (Phocaea).

      Hence, if you find the need to justify some J2 as arriving in historical or proto-historical times, you should look to Asia Minor, where Phocaea, the main Greek colonial power stood and where J2 is more common as well, but Anatolia is also a plausible origin of the Etruscan elites, etc. Other Aegean influences are apparent through all the Late Prehistory: in Neolithic, Chalcolithic and Bronze Age, so no need to overload Classical Greeks with the burden of "paternity" for so many Italians.

      Delete
  5. Interesting for me is that K* appears eeven though they split off L and K2 ( T). It must be very old marker.
    and
    The messapic clearly show the Epirote and Greek markers like the rest of the Greek area in Itlay. The Illyrian fantasy must have been a dream

    ReplyDelete
    Replies
    1. They could well have been Illirized (or Venetized) while retaining most of their ancestral genetics. I have no problem with that. Now they are Latinized, so what?

      Delete
  6. A troubling result is the strong frequency of J2 along Adriatic Sea, specially in Marches and Abruzze.
    Do you have an explanation ? I signal a possible relation with the North-Picene language
    http://fr.wikipedia.org/wiki/Nord-pic%C3%A8ne

    In Apulia, maybe I am wrong but it seems me G2 is more inside (and probably older in the region) and J2 and E-V13 more coastal.

    I did the same remark as you for hg I , but the problem for Sanniti, this represents only 3 tested man.

    Probably, the antique slave trade had an influence but difficult to estimate, probably it will be possible when we will dispose of SNP markers dating from about these times. but outside coastal towns, the servants were more from an local origin (deserted children, debts, kidnapings, illegal abuse by rich land owners, ...) .
    I have my theory about emancipated slaves of imperial and aristocraty families. A lot had never ben real slaves. Simply the poor families of children (specially educated children of greek origin) gave them to the rich patricians by two simultaned acts; A sell of the child and at once an emancipation which continues to link child to the rich family legally. For me this explains a lot of misunterstood things about the emancipated men of the imperial family.

    ReplyDelete
    Replies
    1. It's been known for long that J2 is important in Italy. In principle it'd be yet another Neolithic lineage although as of now it remains to be detected in aDNA of that era.

      "... the problem for Sanniti, this represents only 3 tested man."

      3/30 is still 10%. And it's not a mere singleton, which is much more dubious. It is anyhow also important in Sicily, where other I has not been found at all.

      "Probably, the antique slave trade had an influence but difficult to estimate"...

      I doubt it was an important one. We do not see many exotic lineages, do we? If the slave trade had left such a legacy, we would expect to see significant amounts of anomalous lineages like E-M81 from NW Africa or R1a1-M17 from Eastern Europe. When we look at the most famous plantation province of the Roman era, Sicily, we do see only very little E-M81 (which could also have arrived with Carthaginians) and no R1a1-M17. So I'm rather quite inclined to disregard the slave legacy as insignificant, at least patrilineally. Most Roman slaves had very rough lives, dying early and leaving no descendants. Women may have fared better but I do not see any obvious signature either.

      Delete
    2. PS- Notice that J2(b) is also important in Iberia (particularly in the West, South, Aragon and Majorca), what seems to support the Neolithic connection perceived for the other "Neolithic" lineages (E-V13, G2a and I-M26).

      Delete
  7. So as of 2014, the most recent professional knowledge on ethnic Italian genetics is that the Mitochondrial DNA gene pool is greatly of Palaeolithic origin, and the Y Chromosomal gene pool is greatly of Neolithic to Bronze Age origin, correct? Hence, Phoenician and Ancient Greek colonial contacts, and other incursions like the Germanic invasions, did little to.substantially affect the original populations, am I right? This is how interpret the results of all the latest work by Brisighelli, Boattini, etc. I am an American of 50% Italian descent. Two of my Maternal Great-Grandparents were from Tuscany, one was from Abruzzo, and one was from Campania. Recently, I had genetic testing done by 23andMe, and then Raw Data further analyzed from there. 23andMe's Ancestry Composition tracks back to at least 500 YBP, which in historical terms, of coure is not much, but it was interesting still to see. By SNP analysis, my Maternally-inherited X Chromosome is rendered as about 90% Italian (SNPs specific to Italian populations), nearly 10% Nonspecific Southern European (SNPs possibly shared by populations across Southern Europe), and a tiny fraction was Unassigned (SNPs shared across multiple populations). These later two Results are largely due to conservatism, as the Company is not keen on giving false positives. The other two Reference Populations for Southern Europe are Iberia and the Balkans, and when my Nonspecific Southern European was tested against both of those, I came out as 0.0%. My MtDNA Haplogroup, which I inherited through the Tuscan side, is J, and after later analysis, turns out to be the Subclade J2a1. Not a ton is known about this MtDNA Subclade as of yet apparently, other than it originated in Europe 16,500 YPB, and together with J1c, was one of J's Subclades involved in Paleolithic human settlement. J2a1 is found pretty specifically in central, western, and northern Europe, peaking (barely 5.0%) in the Alps.

    ReplyDelete
    Replies
    1. Ashton you are a little off on the Germanic part and here is why.1 60% of North Italian males are R1B and surprisingly enough 25% of South Italian males are R1B as well.But in Sardinia and Corsica you find more of the Y line G Neolithic haplogroups as well. Just curious you said you are half Italian what is your Ydna and MT DNA haplogroups if you don't mind sharing? I live in the USA as well and my great grandfather was born and lived in Biella in the Northwest part of Italy on the Northern edge of the province of Piedmonte. I took a Y DNA test and am awaiting the results. I will publish them when they come in. My great grandfather also had A+ blood as did grandpa as does my father as do i.Yet mom and grandma (dad's mom had have O+ blood) lol.

      Delete
    2. 0+ is the most common blood group on Earth, relax. We all have it, well not all, but most.

      Delete
  8. This comment has been removed by the author.

    ReplyDelete
    Replies
    1. Absolutely correct (except that you changed "iceman" by "ocean). What's the point with it?

      Delete
  9. Otzi the ICEMAN was Y DNA G2A and K MT DNA but did they determine his blood type? Why you ask? Why not? Here we have a man from the Neolithic period and his two ancient haplogroups so yes Maju it would be very interesting to know that as well especially when it has been theorized that the A B O blood types originated in certain areas of the world that his blood type may either lend creedance to the theory or refute it.Just to give you an example a vast majority of the Greenlandic and north American eskimo/inuit population is B+ blood type and 25% of them carry at least one allele of HLA B-57.

    ReplyDelete
    Replies
    1. No, haploid lineages are (almost) totally unrelated to autosomal genetics, including those genes determining blood type.

      Nowadays, most believe that blood groups may have evolved many times each. Else, group 0 is oldest, A next (and old enough to be shared with Aboriginal Australians) and B a bit more recent maybe (but old enough to be shared by large numbers of Chinese and Indians). So they'd be all at least 80-60 Ka old, unless they have evolved many times.

      In Europe however, blood type B seems quite strictly associated with Indoeuropean expansion from Eastern Europe. The other two types are clearly much older (and common).

      Delete

Please, be reasonably respectful when making comments. I do not tolerate in particular sexism, racism nor homophobia. Personal attacks, manipulation and trolling are also very much unwelcome here.The author reserves the right to delete any abusive comment.

Preliminary comment moderation is... ON (your comment may take some time, maybe days or weeks to appear).