View Full Version : Colonial New England + Caucasus/West Asia?

12-16-2012, 12:57 AM
Members who migrated over from DNA Forums might recall that I've been chasing an elusive elevated "South Asian" component ever since I first submitted my 23andMe results to Doug McDonald about three years ago. Discussion at the time centered on the possibility of Roma ancestry and/or the English colonization of India; none of which really fit the geography or demographics of my family history. With the evolution of various tools since then, this mystery component continues to appear and now seems to point toward a small bit of some kind of Caucasian or Pakistani ancestry.

My mother's most recent results from Doug McDonald:


Most likely fit is 93.1% (+- 1.1%) Europe (all Western Europe)

and 6.9% (+- 1.1%) Mideast (various subcontinents)

The following are possible population sets and their fractions,

most likely at the top

English= 0.910 Adygei= 0.090 or

English= 0.929 Georgian= 0.071 or

English= 0.936 Iranian= 0.064 or

English= 0.933 Turkish= 0.067 or

English= 0.944 Armenian= 0.056

Second most likely fit is 96.8% (+- 1.2%) Europe (all Western Europe)

and 3.2% (+- 1.2%) S. Asia (various subcontinents)

The following are possible population sets and their fractions,

most likely at the top

English= 0.947 Pathan= 0.053 or

English= 0.956 Sindhi= 0.044 or

English= 0.972 N_India= 0.028 or

English= 0.978 S_India= 0.022 or

English= 0.977 Hazara= 0.023 or

English= 0.978 Uygur= 0.022

This clearly is English modified by something more eastern. I looked at several

PCA plots and they all agree on that, but nothing else. It could be anywhere from Holland east.

If in fact Holland or Germany, there’s a lot of it. On the other hand, this sort of “Continental”

English is common (i.e. in me) and often unexplainable. Then there is that tiny 0.1% American,

but it is also weak and therefore has to be considered as in the noise.

My mother's Dodecad Oracle results:

Mixed Mode Population Sharing:

# Primary Population (source) Secondary Population (source) Distance
1 94.3% Cornwall (1000 Genomes) + 5.7% Kalash @ 0.44
2 96.1% Dutch (Dodecad) + 3.9% Pathan (HGDP) @ 0.54
3 96.4% Dutch (Dodecad) + 3.6% Sindhi (HGDP) @ 0.54
4 94.2% Kent (1000 Genomes) + 5.8% Kalash @ 0.56
5 96.2% Dutch (Dodecad) + 3.8% Burusho (HGDP) @ 0.66
6 96.6% Dutch (Dodecad) + 3.4% Kalash @ 0.67
7 93.3% British_Isles (Dodecad) + 6.7% Brahui (HGDP) @ 0.68
8 93.2% British_Isles (Dodecad) + 6.8% Balochi (HGDP) @ 0.72
9 95.7% German (Dodecad) + 4.3% Sindhi (HGDP) @ 0.72
10 95.4% German (Dodecad) + 4.6% Pathan (HGDP) @ 0.74

Dodecad Oracle using K12b

Mixed Mode Population Sharing:

# Primary Population (source) Secondary Population (source) Distance
1 80.8% Mixed_Germanic (Dodecad) + 19.2% Hungarians (Behar) @ 1.36
2 90.1% Mixed_Germanic (Dodecad) + 9.9% Ukranians (Yunusbayev) @ 1.54
3 62.2% Mixed_Germanic (Dodecad) + 37.8% German (Dodecad) @ 1.57
4 91.3% Mixed_Germanic (Dodecad) + 8.7% Mordovians (Yunusbayev) @ 1.69
5 93.2% Dutch (Dodecad) + 6.8% Chechens (Yunusbayev) @ 1.72
6 91.6% Mixed_Germanic (Dodecad) + 8.4% Russian_B (Behar) @ 1.75
7 90.5% Mixed_Germanic (Dodecad) + 9.5% Mixed_Slav (Dodecad) @ 1.77
8 93.7% Dutch (Dodecad) + 6.3% North_Ossetians (Yunusbayev) @ 1.81
9 93.6% Dutch (Dodecad) + 6.4% Balkars (Yunusbayev) @ 1.82
10 93.6% Dutch (Dodecad) + 6.4% Adygei (HGDP)

My mother and her two sisters also have several Georgian, Armenian, and Turkish matches on Gedmatch; however they have no matches from Pakistan or India that I'm aware of.

Based on how complete our family genealogy is, whatever it is, there is NOT a lot of it. There is nothing non-Isles within 5-6 generations, barring some kind of strange and implausible NPE. The most plausible scenarios would involve >90% English or Dutch with <10% something else.

My mother and aunts all get a flatline zero Ashkenazi percentage on Ancestry Finder (no 4gp Ashkenazi matches with the gates wide open), however they have some unexpected matches from Brazil, Italy, Portugal, the Baltics, and Eastern Europe which makes me wonder about Sephardic ancestry.

Their documented ancestry is 1/4 English, from 1840s/50s immigration (Northamptonshire and Cornwall), and 3/4 colonial American, nearly all from New England. The entire family was in the American upper Midwest by 1850 where even an NPE is unlikely to result in anything too exotic. We have photos of all branches back to the late 1800s and they all look typically NW European, with the exception of one of my great-grandfathers, who looks a bit darker and more Mediterranean than your typical Englishman. He was born in 1884 in northern Wisconsin. His father , surnamed Stephens, was a logger of New England ancestry whose family migrated to Illinois in the 1840s and Wisconsin in the 1850s. His mother's family, surnamed Ide, was from upstate New York, with migration to Michigan in the 1840s. The Ide branch originally came from colonial Rhode Island and New Netherlands, and includes mostly Yankee with a bit of Dutch and Huguenot.

The Ide branch of the family has several brick-walls and gaps around 1750, which I think is recent enough to account for this admixture, if I could only figure out how Georgians ended up in colonial Rhode Island. My best guess is a sailor or escaped slave in a New England port. Evidently there were "Turks" in colonial Rhode Island, as they're mentioned in historical documents; but I can't find any more specific reference to them.