23andme me results of Sephardi Jew

[Sp_loa]

North African Sephardim.

Known ancestry:

3/4 from Former Spain (Spanish Morocco):
1/2 from Tangier
1/4 from Larache

and
1/4 from Casablanca


Autosomal:

https://i.ibb.co/0q13TsL/Screen-Shot-2019-03-27-at-2-54-20-PM.png


Regional Matches:


Spain:

1. Aragon
2. Galicia

Italy:
1. Calabria
2. Sicily

Algeria:
1. Oran

Morocco:
1. Fez-Meknes
2. Casablanca Settat
3. Rabat-Sale

Mtdna: U2e1 (found in central, western and northern Europe)


Very specific results because of her recent Spanish roots (Tangier-Tetouan region hosted many conversos from Iberia and italy).

Mix of South Italian-Northern Spaniard-Levantine and some Northern Berber. Very European, Very interesting.

[Kamal900]

Very nice results. Is she from Israel?

[Sp_loa]

Very nice results. Is she from Israel?

I'm not sure.
She is the first Moroccan Jew I saw with match for Spain. Galicia and Aragon make sense to me. I have 3rd-4th cousins Puerto-ricans (gentiles) cousins and we narrowed down our common ancestor to a family (christian) that lived in Galicia (NW Spain) until the early 18th century. Explains why some people in my family look very distinctively NW Iberian. I wish my grandmother would have gotten such specific breakdowns.

[Sp_loa]

If you guys so my other thread, You must be really shocked (as I am) that Moroccan Jews can be anything from 70%+ MENA to 60%+ European depends on the regions.....
Tangier-Tetouan-Al hoceima was the most European region. I think the Jews from there are North shifted to all types of Sephardim.

[ModernMaskil]

How common is something like this? Altogether with all the Moroccan matches you've had are most majority Euro or majority MENA by 23andme's classification?

[Sp_loa]

How common is something like this? Altogether with all the Moroccan matches you've had are most majority Euro or majority MENA by 23andme's classification?

Again, Moroccan Jew family won't be just randomly 60% Euro and then their next door neighbors will be 75% MENA. It's regional due to historical reasons- Sephardi Jews settled mainly in the big coastal cities, and the purest Sephardic region (they kept their Spanish identity and language to this day) was in Tangier-Tetouan-Al hoceima (that ironically became Spain again for a few decades). More than this- some families of Conversos (that were already "recently" Iberian mixed, and very European genetically) settled in Northern Morocco in order to return to Judaism.

On the other hand- less Sephardim reached to the South, so Southern you go the Jewish population resembles more the pre-Sephardic Moroccan Jewish population- mix of mainly Levantine and Berber. AKA the European decreases.

BTW Those are my grandma relatives (hence obviously mine too but I have other sides...). Her matches are diverse as you saw. Most of them are around 45%-50% Euro, in par with the genetic admix of Jews from Fez-Meknes , where most of her 23andme relatives are from.

[Kamal900]

If you guys so my other thread, You must be really shocked (as I am) that Moroccan Jews can be anything from 70%+ MENA to 60%+ European depends on the regions.....
Tangier-Tetouan-Al hoceima was the most European region. I think the Jews from there are North shifted to all types of Sephardim.

Can't say I'm that shocked considering that western Jewish peoples like the Ashkenazi and the Sephardim are around 50 to 55% Europeans genetically on average.

[Sp_loa]

Can't say I'm that shocked considering that western Jewish peoples like the Ashkenazi and the Sephardim are around 50 to 55% Europeans genetically on average.

Southernmost Moroccan Jews are 80%+ MENA

[Kamal900]

Southern most Moroccan Jews are 80%+ MENA

I guess it make sense. I think it all depends on the region. The same's true for Palestinians for example. My family come from the very north of Israel.

[Morena]

nice results

[ModernMaskil]

Again, Moroccan Jew family won't be just randomly 60% Euro and then their next door neighbors will be 75% MENA. It's regional due to historical reasons- Sephardi Jews settled mainly in the big coastal cities, and the purest Sephardic region (they kept their Spanish identity and language to this day) was in Tangier-Tetouan-Al hoceima (that ironically became Spain again for a few decades). More than this- some families of Conversos (that were already "recently" Iberian mixed, and very European genetically) settled in Northern Morocco in order to return to Judaism.

On the other hand- less Sephardim reached to the South, so Southern you go the Jewish population resembles more the pre-Sephardic Moroccan Jewish population- mix of mainly Levantine and Berber. AKA the European decreases.

BTW Those are my grandma relatives (hence obviously mine too but I have other sides...). Her matches are diverse as you saw. Most of them are around 45%-50% Euro, in par with the genetic admix of Jews from Fez-Meknes , where most of her 23andme relatives are from.

Thanks.
One thing I noticed is there is a lot of "broadly" this and that. 35% in fact. Frustrating.

[Sp_loa]

Thanks.
One thing I noticed is there is a lot of "broadly" this and that. 35% in fact. Frustrating.

It's still better than Guessing. AncestryDNA in many cases just guesses and then you get unreasonable results like mexican scoring 40% Portuguese and 5% Spanish (when Mexico was colonized by Spaniards) just because both populations are similar. I prefer "Broadly" over wrong guesses.

[Leto]

Any chance to see her GEDmatch results?

[Sp_loa]

Any chance to see her GEDmatch results?

I'm trying to get it too but I don't think she ever uploaded her raw data to gedmatch. If she ever does I'll update you.

[Sp_loa]

up

[Pine]

The regions she matches in Italy are known for converso blood. The same may be true for Spain, but I'm not as familiar with Spain. Her "European" % is inflated. As for calculating her "Berber", you need to keep in mind that some Levantines do get around the same % North African on 23AndMe and every other Moroccan Muslim I meet, tells me he descends from some Berber tribe that used to be Jewish.

[Sp_loa]

The regions she matches in Italy are known for converso blood. The same may be true for Spain, but I'm not as familiar with Spain. Her "European" % is inflated. As for calculating her "Berber", you need to keep in mind that some Levantines do get around the same % North African on 23AndMe and every other Moroccan Muslim I meet, tells me he descends from some Berber tribe that used to be Jewish.

Her European percentage isn’t really off for Jews from Northern Morocco especially when we are talking about South Italy+Spain. 23andme seem to be reasonable with European percentages for everyone- Turkish Jews score around 30-40% European and Jews from central morocco also that amount.

About North Africa- maybe, but some Jews are overwhelmingly Berber. 23andme should separate Berber from Arabia.

My grandmother Scored 50% and it matches her Gedmatch oracles (the regions also).

[Pine]

Her European percentage isn’t really off for Jews from Northern Morocco especially when we are talking about South Italy+Spain. 23andme seem to be reasonable with European percentages for everyone- Turkish Jews score around 30-40% European and Jews from central morocco also that amount.

About North Africa- maybe, but some Jews are overwhelmingly Berber. 23andme should separate Berber from Arabia.

My grandmother Scored 50% and it matches her Gedmatch oracles (the regions also).

What's your reference source for these %'s?

How do you judge a gedmatch oracle to be accurate?

I do it from an algebraic intuition. If I see a great reduction in distance when the amount of dimensions in the model is concerned relative to the amount of vectors selected to fit the individual being tested, then I consider it a "good" fit.

Example (My K13):

# Population (source) Distance
1 Sephardic_Jewish 5.02

# Primary Population (source) Secondary Population (source) Distance
1 64.9% West_Sicilian + 35.1% Lebanese_Druze @ 2.53

Note the presence of southern Italian, which you argue is the likelier contender for south euro admixture into western Jews. The additional choice of a vector barely cuts the distance in half.

Whereas,

My MDLP World (12 dimensions, thus comparable to the Eurogenes K13 example above):


# Population (source) Distance
1 Greek_Cretan 5.48


# Primary Population (source) Secondary Population (source) Distance
1 57.9% Lebanese + 42.1% Italian-North @ 1.17

Notice the factor by which the distance is reduced! The factor is 4-5x.

I'm not being selective here either. I'm pretty sure this is the best fit I get on gedmatch, aside from maybe some pre-historic ones.

[Sp_loa]

What's your reference source for these %'s?

How do you judge a gedmatch oracle to be accurate?

I do it from an algebraic intuition. If I see a great reduction in distance when the amount of dimensions in the model is concerned relative to the amount of vectors selected to fit the individual being tested, then I consider it a "good" fit.

Example (My K13):

# Population (source) Distance
1 Sephardic_Jewish 5.02

# Primary Population (source) Secondary Population (source) Distance
1 64.9% West_Sicilian + 35.1% Lebanese_Druze @ 2.53

Note the presence of southern Italian, which you argue is the likelier contender for south euro admixture into western Jews. The additional choice of a vector barely cuts the distance in half.

Whereas,

My MDLP World (12 dimensions, thus comparable to the Eurogenes K13 example above):


# Population (source) Distance
1 Greek_Cretan 5.48


# Primary Population (source) Secondary Population (source) Distance
1 57.9% Lebanese + 42.1% Italian-North @ 1.17

Notice the factor by which the distance is reduced! The factor is 4-5x.

I'm not being selective here either. I'm pretty sure this is the best fit I get on gedmatch, aside from maybe some pre-historic ones.

I’m not sure how accurate it is to rely on distances as I saw sicilians getting 78% Sephardic+22% West german as best distance (less than 1) and obviously it isn’t true.

Anyway all researchers today agree that Ashkenazim are around 15-25% East European. If they are also 35% North Italian and 50% Levantine, they should have plotted Northern to Sicilians.

I find 23andme very accurate when it comes to European ancestry so I trust them.

There is no way proving or disproving my claim. To this day researchers aren’t sure what is the origin of Southern European admix of Jews.
If you want to believe Ashkenazim are just half Lombard or Aostans and half Levantine, (ignoring their IBD sharing with Slavs).. be my guest


BTW on my k13 best oracle is 96% Sephardi Jew+ 4% Bantu North (2.58) and only the 17th oracle is 86.6% Sephardi Jew+ 13.4% Moroccan (3.36).

You tell me which one is more reasonable.

[waam]

I think GEDmatch isn't to be taken too literally (I score 90+% Lebanese in the oracles which I'm obviously not) but 23andme isn't too well refined either imo. until they break down the MENA region into sub categories it should be taken with a grain of salt.

[Sp_loa]

I think GEDmatch isn't to be taken too literally (I score 90+% Lebanese in the oracles which I'm obviously not) but 23andme isn't too well refined either imo. until they break down the MENA region into sub categories it should be taken with a grain of salt.

About 23andme
For their MENA I agree, but their European reference, especially Southern european, is very good.

[Pine]

I’m not sure how accurate it is to rely on distances as I saw sicilians getting 78% Sephardic+22% West german as best distance (less than 1) and obviously it isn’t true.

So what? Was it their best distance reduction from single mode to mixed mode of all gedmatch models? Because that's the metric I proposed using. Also, Around 9,000 Jews converted in order to remain in Sicily.[9]
(from wiki). 9k wasn't a tiny number relative to the total at the time.


Anyway all researchers today agree that Ashkenazim are around 15-25% East European.

All? I'm pretty sure you're alluding to a single study about "the time and place of european admixture in Ashkenazi Jews" or some shit like that, where a number of assumptions went into that estimate. NMonte calculators don't give a number nearly as high.


If they are also 35% North Italian and 50% Levantine, they should have plotted Northern to Sicilians.

Individual Jews do - look at the 2019 PCA map on here. And your conclusion isn't necessarily true. Endogamy would've still went on after the supposed east euro admixture event, whereby the east euro segments may have been randomly inherited at lower %'s. Also, I've seen PCAs where Ashkenazim do plot north of Sicilians. Regadless, we know I have next to 0 east euro admix (as tested genetically) and my best fit , using the aforementioned criteria, is in fact North Italian + Lebanese. The only way 25% East Euro admix could've been true is through a large % of it inherited from East Knaanic Jews, as a large scale conversion of east euro women during Ashkenazi migration into eastern Europe would've been noted in Rabbinical literature. Rabbis were already writing plenty at the time.


I find 23andme very accurate when it comes to European ancestry so I trust them.

There is no way proving or disproving my claim. To this day researchers aren’t sure what is the origin of Southern European admix of Jews.
If you want to believe Ashkenazim are just half Lombard or Aostans and half Levantine, (ignoring their IBD sharing with Slavs).. be my guest


BTW on my k13 best oracle is 96% Sephardi Jew+ 4% Bantu North (2.58) and only the 17th oracle is 86.6% Sephardi Jew+ 13.4% Moroccan (3.36).

You tell me which one is more reasonable.

You missed my point again. Is that mixed mode estimate a) a massive reduction from the single population distance you get and b) is it the best distance reduction you get on gedmatch? I bet neither are true.

[Leto]

What's your reference source for these %'s?

How do you judge a gedmatch oracle to be accurate?

I do it from an algebraic intuition. If I see a great reduction in distance when the amount of dimensions in the model is concerned relative to the amount of vectors selected to fit the individual being tested, then I consider it a "good" fit.

Example (My K13):

# Population (source) Distance
1 Sephardic_Jewish 5.02

# Primary Population (source) Secondary Population (source) Distance
1 64.9% West_Sicilian + 35.1% Lebanese_Druze @ 2.53

Note the presence of southern Italian, which you argue is the likelier contender for south euro admixture into western Jews. The additional choice of a vector barely cuts the distance in half.

Whereas,

My MDLP World (12 dimensions, thus comparable to the Eurogenes K13 example above):


# Population (source) Distance
1 Greek_Cretan 5.48


# Primary Population (source) Secondary Population (source) Distance
1 57.9% Lebanese + 42.1% Italian-North @ 1.17

Notice the factor by which the distance is reduced! The factor is 4-5x.

I'm not being selective here either. I'm pretty sure this is the best fit I get on gedmatch, aside from maybe some pre-historic ones.
What do you get on Dodecad K12b? I'm also interested in your North European percentage.

[Pine]

What do you get on Dodecad K12b? I'm also interested in your North European percentage.

Admix Results (sorted):

# Population Percent
1 Caucasus 36.05
2 Atlantic_Med 23.11
3 Southwest_Asian 14.55
4 North_European 10.84
5 Gedrosia 6.86
6 Northwest_African 5.36
7 Southeast_Asian 2.54
8 East_African 0.58
9 East_Asian 0.12

Single Population Sharing:

# Population (source) Distance
1 Ashkenazy_Jews (Behar) 6.5
2 Ashkenazi (Dodecad) 6.55
3 Sephardic_Jews (Behar) 6.59
4 S_Italian_Sicilian (Dodecad) 7.8
5 Sicilian (Dodecad) 7.9
6 Morocco_Jews (Behar) 9.23
7 Greek (Dodecad) 12.3
8 Cypriots (Behar) 14.31
9 C_Italian (Dodecad) 14.65
10 Lebanese (Behar) 16.44
11 Turkish (Dodecad) 16.6
12 O_Italian (Dodecad) 17.47
13 Tuscan (HGDP) 18.71
14 Turks (Behar) 18.75
15 Syrians (Behar) 19.65
16 TSI30 (Metspalu) 19.81
17 Druze (HGDP) 20.15
18 Jordanians (Behar) 20.48
19 Palestinian (HGDP) 22.14
20 Uzbekistan_Jews (Behar) 22.32

Mixed Mode Population Sharing:

# Primary Population (source) Secondary Population (source) Distance
1 53% C_Italian (Dodecad) + 47% Lebanese (Behar) @ 3.46
2 91% Sephardic_Jews (Behar) + 9% Mordovians (Yunusbayev) @ 3.57
3 91.8% Ashkenazi (Dodecad) + 8.2% Makrani (HGDP) @ 3.7
4 92.8% Ashkenazi (Dodecad) + 7.2% Brahui (HGDP) @ 3.76
5 91.4% Sephardic_Jews (Behar) + 8.6% Russian_B (Behar) @ 3.77
6 92.3% Ashkenazi (Dodecad) + 7.7% Balochi (HGDP) @ 3.79
7 91.6% Sephardic_Jews (Behar) + 8.4% Russian (HGDP) @ 3.79
8 84.3% Ashkenazi (Dodecad) + 15.7% Iranians (Behar) @ 3.85
9 92.3% Sephardic_Jews (Behar) + 7.7% Lithuanian (Dodecad) @ 3.89
10 69.9% Sicilian (Dodecad) + 30.1% Lebanese (Behar) @ 3.89
11 92.7% Sephardic_Jews (Behar) + 7.3% Lithuanians (Behar) @ 3.9
12 90.8% Sephardic_Jews (Behar) + 9.2% Ukranians (Yunusbayev) @ 3.91
13 91.2% Sephardic_Jews (Behar) + 8.8% Mixed_Slav (Dodecad) @ 3.91
14 91.7% Sephardic_Jews (Behar) + 8.3% Belorussian (Behar) @ 3.94
15 92.9% Sephardic_Jews (Behar) + 7.1% Finnish (Dodecad) @ 3.98
16 92.8% Sephardic_Jews (Behar) + 7.2% FIN30 (1000Genomes) @ 3.99
17 91.1% Sephardic_Jews (Behar) + 8.9% Chuvashs (Behar) @ 4.01
18 54.9% Lebanese (Behar) + 45.1% TSI30 (Metspalu) @ 4.04
19 83.8% Ashkenazi (Dodecad) + 16.2% Kurd (Dodecad) @ 4.08
20 53.4% Lebanese (Behar) + 46.6% Tuscan (HGDP) @ 4.11

[Leto]

Admix Results (sorted):

# Population Percent
1 Caucasus 36.05
2 Atlantic_Med 23.11
3 Southwest_Asian 14.55
4 North_European 10.84
5 Gedrosia 6.86
6 Northwest_African 5.36
7 Southeast_Asian 2.54
8 East_African 0.58
9 East_Asian 0.12

Single Population Sharing:

# Population (source) Distance
1 Ashkenazy_Jews (Behar) 6.5
2 Ashkenazi (Dodecad) 6.55
3 Sephardic_Jews (Behar) 6.59
4 S_Italian_Sicilian (Dodecad) 7.8
5 Sicilian (Dodecad) 7.9
6 Morocco_Jews (Behar) 9.23
7 Greek (Dodecad) 12.3
8 Cypriots (Behar) 14.31
9 C_Italian (Dodecad) 14.65
10 Lebanese (Behar) 16.44
11 Turkish (Dodecad) 16.6
12 O_Italian (Dodecad) 17.47
13 Tuscan (HGDP) 18.71
14 Turks (Behar) 18.75
15 Syrians (Behar) 19.65
16 TSI30 (Metspalu) 19.81
17 Druze (HGDP) 20.15
18 Jordanians (Behar) 20.48
19 Palestinian (HGDP) 22.14
20 Uzbekistan_Jews (Behar) 22.32

Mixed Mode Population Sharing:

# Primary Population (source) Secondary Population (source) Distance
1 53% C_Italian (Dodecad) + 47% Lebanese (Behar) @ 3.46
2 91% Sephardic_Jews (Behar) + 9% Mordovians (Yunusbayev) @ 3.57
3 91.8% Ashkenazi (Dodecad) + 8.2% Makrani (HGDP) @ 3.7
4 92.8% Ashkenazi (Dodecad) + 7.2% Brahui (HGDP) @ 3.76
5 91.4% Sephardic_Jews (Behar) + 8.6% Russian_B (Behar) @ 3.77
6 92.3% Ashkenazi (Dodecad) + 7.7% Balochi (HGDP) @ 3.79
7 91.6% Sephardic_Jews (Behar) + 8.4% Russian (HGDP) @ 3.79
8 84.3% Ashkenazi (Dodecad) + 15.7% Iranians (Behar) @ 3.85
9 92.3% Sephardic_Jews (Behar) + 7.7% Lithuanian (Dodecad) @ 3.89
10 69.9% Sicilian (Dodecad) + 30.1% Lebanese (Behar) @ 3.89
11 92.7% Sephardic_Jews (Behar) + 7.3% Lithuanians (Behar) @ 3.9
12 90.8% Sephardic_Jews (Behar) + 9.2% Ukranians (Yunusbayev) @ 3.91
13 91.2% Sephardic_Jews (Behar) + 8.8% Mixed_Slav (Dodecad) @ 3.91
14 91.7% Sephardic_Jews (Behar) + 8.3% Belorussian (Behar) @ 3.94
15 92.9% Sephardic_Jews (Behar) + 7.1% Finnish (Dodecad) @ 3.98
16 92.8% Sephardic_Jews (Behar) + 7.2% FIN30 (1000Genomes) @ 3.99
17 91.1% Sephardic_Jews (Behar) + 8.9% Chuvashs (Behar) @ 4.01
18 54.9% Lebanese (Behar) + 45.1% TSI30 (Metspalu) @ 4.04
19 83.8% Ashkenazi (Dodecad) + 16.2% Kurd (Dodecad) @ 4.08
20 53.4% Lebanese (Behar) + 46.6% Tuscan (HGDP) @ 4.11
The Asian percentage is too large to be noise, I believe. Surprisingly you're still 10% North European.

53% C_Italian (Dodecad) + 47% Lebanese (Behar) @ 3.46

Nice fit.

[Pine]

The Asian percentage is too large to be noise, I believe. Surprisingly you're still 10% North European.

53% C_Italian (Dodecad) + 47% Lebanese (Behar) @ 3.46

Nice fit.

The Ashkenazi mean is 13%. Trying to figure out if it's random or through non-Ashkenazi admix. The Southeast Asian is large. I'm trying to figure out if it's miscategorized northeast asian + siberian or something else. However, it's always southeast asian for me on all calcs. I never get Siberian, unlike other Ashkenazim. Why do I keep getting Southeast Asian instead of Northeast Asian? The fit barely reduces the single population distance though.

[Leto]

Don't wanna start a new thread, people usually ignore my posting anyway. Just found this girl, Australian father (Anglo-Celtic), Jewish mother (Ashkenazi from Poland and Sephardic from Turkey)


https://www.youtube.com/watch?v=s7OPPal292Y

She looks very white, I was surprised she's not 100% European.

[Pine]

Don't wanna start a new thread, people usually ignore my posting anyway. Just found this girl, Australian father (Anglo-Celtic), Jewish mother (Ashkenazi from Poland and Sephardic from Turkey)


https://www.youtube.com/watch?v=s7OPPal292Y

She looks very white, I was surprised she's not 100% European.

Has East Med/West Asian (whatever you want to call it) eyes. Arguebly looks more Jewish than I do and I am autosomally the most semitic Ashkenazi as far as I know. How many Jews have you seen before? Not saying that she looks Jewish or I'd guess her as a Jew. Some Jewish ancestry isn't a surprise here though.

[Leto]

Has East Med/West Asian (whatever you want to call it) eyes. Arguebly looks more Jewish than I do and I am autosomally the most semitic Ashkenazi as far as I know. How many Jews have you seen before? Not saying that she looks Jewish or I'd guess her as a Jew. Some Jewish ancestry isn't a surprise here though.
I've seen enough. If she looks Jewish, then looking Jewish has little to do with looking Semitic/Levantine. She's 1/4 Middle Eastern at most! A real Levantine person would laugh at that.

[Pine]

I've seen enough. If she looks Jewish, then looking Jewish has little to do with looking Semitic/Levantine. She's 1/4 Middle Eastern at most! A real Levantine person would laugh at that.

I never said she looks Jewish. I said her eyes don't make the revelation of some Jewish ancestry a "surprise". In fact, I recall seeing eyes like that on Sephardim more than on Ashkenazim. If asked to guess her ancestry, I would've guessed Northwest Euro with something like Greek down the line. No levantine would laugh at such an assertion.