CMH Question: I was but am I still?

[josh w.]

Vinnie, just to repeat Jim's point about DYS 388. DYS 388=17 is related to the spread of Islam. A 388 value of 16 could suggest Jewish origins. However, remember that some Jewish and I&P Arab haplotypes overlap, at least at the six marker level.

[vinnie]

My 388 = 16, but what do you mean "I&P"?

[sundodger]

My 388 = 16, but what do you mean "I&P"?

I think he means Israeli and Palestinian.

[vinnie]

Thank you!

[cacio]

I briefly looked at a couple of papers, and here are some J percentages in various places in Italy. Note, samples tend to be small, so you'll see that the results vary between studies.

First column J1, second column J2

From Capelli 2006:
Central Tuscany 7 17
Elba Island 0 8
Central Marche 5 36
Appennine Marche 0 22
Tuscany Latium border 0 19
NE Latium 0 15
S Latium 0 25
NW Apulia 2 17
W Campania 5 17
S Apulia 1 24
W Calabria 2 35

From Di Giacomo 2003:
Verona 0 27
Garfagnana 0 9
Genoa 3 7
L'Aquila 3 31
Pescara 15 15
Avezzano 3 17
Benevento 6 20
Cilento 6 21
Foggia 0 44
N Gargano 17 20
Casarano 0 25
Brindisi 2 20
Altamura 0 12
Matera 0 12
Paola 11 31
Reggio C 9 27

From Capelli 2005:
E Sicily 7 28
SW Sicily 0 27
NW Sicily 7 11
S Italy 4.5 16
Sardinia 5 9

From Semino 2004:
NC Italy 0 27
Calabria 1 21
Puglia 2 29
Sicily 7 16
Sardinia 3 10

cacio

[vineviz]

I briefly looked at a couple of papers, and here are some J percentages in various places in Italy. Note, samples tend to be small, so you'll see that the results vary between studies.

Thanks for compiling that data.

Here's a visual summary of the frequency of haplogroup J in Italy, collected from various sources.

http://vizachero.com/images/HapJ.png

[vinnie]

[QUOTE=cacio]I briefly looked at a couple of papers, and here are some J percentages in various places in Italy. Note, samples tend to be small, so you'll see that the results vary between studies.

Cacio, thanks much for sending this - you didn't have to do this much work! My grandfather's town, Villavallelonga, is literally the last bus stop in the Appenines south of Avezzano, which is in l'Aquila/Abruzzo.

[vinnie]

Thanks for compiling that data.

Here's a visual summary of the frequency of haplogroup J in Italy, collected from various sources.

http://vizachero.com/images/HapJ.png

Thank you for the map!

[cacio]

Vince:

very informative map. The cline appears very clearly in this graphical representation. In the raw data I wrote down, it was less clear.

Some comments from steering at your graph and at the data. The overall pattern seems to be dictated by J2, as it is much more frequent. A quick glance at the data does indeed suggest some decrease in J2 as you move N. But what is interesting is that J2 remains high (>20%) into Tuscany and the Marche. The big drop happens really with the Appennine (I suspect the map smooths out the breaks). Equally interesting, Sicily is not the max. Calabria is. Sicily is around the levels of the rest of S Italy. (This reminds me of an old map of Cavalli Sforza, based not on Y/mtdna, but on other markers. He found the first principal component was based in Reggio Calabria and Messina, not in Sicily).

With J1 the situation is much less clear. Partly, this is due to the smaller frequencies, at these level, one is bound to find lots of variation. However, from looking at the data, I'd say that Sicily is here clearly at the top, with a consistent 7% or so. After that, I don't really detect any clinal pattern btw S and C Italy. While this may be due to the small size, there may be more to it. The higher fraction in Sicily may be easily attributed to the Arabs in the Middle Age. However, if they had been the main source of J1 in Italy, I'd have expected to see more of a cline. Instead, we don't, which I would take as evidence that J1 in S and C Italy has other sources. Neolithic, Roman trades and slaves (including Jews). I wonder if anything is known about the Italian Jews - at the haplotype level, I mean.

cacio

[cacio]

Vinnie:

since you mention. Di Giacomo 2003:
Avezzano: sample size 29 (so pretty small)
R1b 41 R1a 7 E 3.5 G2 7 I 10 J2 17 J1 3 Other (ie K2, L etc.) 10

L'Aquila: sample size 35 (also small)
R1b 26 R1a 6 E 11 G2 6 I 8 J2 31 J1 3 Other (ie K2, L etc.) 8

cacio

[vineviz]

I, too, think that J1 probably has an ancient source in Italy. In any event J1, J2a, and J2b clearly have different stories to tell.

Here are three more maps: for J1, J2a, and J2b. These depend on predictions based on minimal haplotypes, so they should be viewed with caution. Also, the scales are different in each map (i.e. bright red in one is not the same frequency as bright red in the others).

http://www.vizachero.com/images/HapJ1.png

http://www.vizachero.com/images/HapJ2a.png

http://www.vizachero.com/images/HapJ2b.png

[vinnie]

I wonder if anything is known about the Italian Jews - at the haplotype level, I mean.

cacio[/QUOTE]

Is there anything published on the haplogroups of modern Italian Jews?

[cacio]

Vince:

nice maps. I am ignorant of J - so I don't know if there is any consensus about the origin and meaning of J2a vs J2b. Anyway, J2a seems to have the clearest cline and what one would expect from an Anatolian migration. My conjecture about J1 being highest in Sicily is not confirmed by the data- perhaps a further sign of the low genetic impact of the Arabs? J2b has smaller frequencies, so as usual there may be random variability. Still, it's hard to dismiss the fact that it goes all the way up to the NW - you would almost take it as an autochtonous haplogroup...

Your maps are telling very interesting stories.

cacio

[MMaddi]

Vince:

very informative map. The cline appears very clearly in this graphical representation. In the raw data I wrote down, it was less clear.

Some comments from steering at your graph and at the data. The overall pattern seems to be dictated by J2, as it is much more frequent. A quick glance at the data does indeed suggest some decrease in J2 as you move N. But what is interesting is that J2 remains high (>20%) into Tuscany and the Marche. The big drop happens really with the Appennine (I suspect the map smooths out the breaks). Equally interesting, Sicily is not the max. Calabria is. Sicily is around the levels of the rest of S Italy. (This reminds me of an old map of Cavalli Sforza, based not on Y/mtdna, but on other markers. He found the first principal component was based in Reggio Calabria and Messina, not in Sicily).

With J1 the situation is much less clear. Partly, this is due to the smaller frequencies, at these level, one is bound to find lots of variation. However, from looking at the data, I'd say that Sicily is here clearly at the top, with a consistent 7% or so. After that, I don't really detect any clinal pattern btw S and C Italy. While this may be due to the small size, there may be more to it. The higher fraction in Sicily may be easily attributed to the Arabs in the Middle Age. However, if they had been the main source of J1 in Italy, I'd have expected to see more of a cline. Instead, we don't, which I would take as evidence that J1 in S and C Italy has other sources. Neolithic, Roman trades and slaves (including Jews). I wonder if anything is known about the Italian Jews - at the haplotype level, I mean.

cacio

Perhaps it's useful to look at the breakdown we have in the Sicily Project. We have 112 yDNA results so far, but there are 4 pairs of results from the same paternal line. So essentially we have 108 distinct yDNA results.

Of those 108, 32 are J (either J1 or J2), representing about 30% of the total. I wish I could give a breakdown by region of Sicily, say eastern vs. western, but I don't have those figures. J is further broken down by 5 J1 and 27 J2. The J1 figure is 4.6% of our total, a little less than the 7% figure you cite above. Perhaps that's due to not enough results yet in the project.

Regarding J1, I think Jim Honeychuck had earlier mentioned in this thread that 2 of our J1's have DYS388=13, which is noticably lower than all J2's and most J1's. According to Bonnie Schrack of the J haplogroup project, this matches a finding in one of Cinnioglu's studies of Turkish subjects. So Bonnie theorizes that a J1 in Europe with DYS388=13 may be a descendant of a very early Neolithic migration from Anatolia west into the Mediterranean.

Mike Maddi

[MMaddi]

Vince:

nice maps. I am ignorant of J - so I don't know if there is any consensus about the origin and meaning of J2a vs J2b. Anyway, J2a seems to have the clearest cline and what one would expect from an Anatolian migration. My conjecture about J1 being highest in Sicily is not confirmed by the data- perhaps a further sign of the low genetic impact of the Arabs? J2b has smaller frequencies, so as usual there may be random variability. Still, it's hard to dismiss the fact that it goes all the way up to the NW - you would almost take it as an autochtonous haplogroup...

Your maps are telling very interesting stories.

cacio

As I noted in my previous posting, of the 5 J1's in the project, Bonnie Schrack believes that 2 of them represent a very early Anatolian migration into the Mediterranean. So right there, that takes away 40% of our J1's from a possible Arab contribution to Sicily's genetic heritage.

I think one place to look for possible Arab contribution to Sicily's gene pool is with E3b and G2.

In the case of E3b in the Sicily Project, we have 18 paternal lines from that haplogroup. Two of them have been SNP-tested and are M81+. They both have DYS385a-b=13-14, much different from most of the other E3b's. There are another two E3b's (non-SNP-tested) who have DYS385a-b=13-14, so I believe that they are probably also M81+. This would indicate that 4 of the 18 E3b paternal lines are M81+. Of course, M81+ is found in Berber populations, not Arab. But if you are looking for some sign of lasting contribution to Sicily from the rulers and colonists during the Muslim period in the Middle Ages, that would fit what you're looking for. I believe most of the Muslims in Sicily during that period were Berber, not Arab.

Also, I've been surprised by the high level of G/G2 in our yDNA results. We have 11 of them. My understanding is that G2 is found at relatively high levels in Arab populations. Someone please correct if I'm wrong about that or if there are other possibilities for deep ancestry of a G/G2 with recent Sicilian ancestry.

Mike