Search Result

[NYMark]

For me, at least, My Origins pretty poorly reflects my recent ancestry.

As I've mentioned on a number of threads, my Ashkenazi percentage, which recently inched up by another point to 57% is almost undoubtedly a significant overestimate. There is no DNA evidence of Ashkenazi ancestry...

[NYMark]

My half brother has a zero score for East European and 1.66% for South Baltic. His mixed mode sharing is 100% Ashkenazi.

Maternal ancestry is all from within the borders of modern Germany, all 19th century immigrants. Paternal is unknown.

As I've mentioned upthread, my paternal...

[NYMark]

As mentioned, I get 56% Ashkenazi on MyOrigins. I'm no longer on Ancestry, but I believe that number was about the same.

On 23andMe, I'm 49% in Standard and 50.9% in speculative. My general sense is that 23andMe in Standard mode is the most accurate, not just for me but across the board...

[NYMark]

One thing that may provide a clue is the number of matches and the number who seem to have Ashkenazi ancestry based on surname. For example, I have a match who is 6% Ashkenazi per My Origins. She has 31 pages of matches in common with me (50%) and 41 pages of matches in common with my half-brother (100%)....

[NYMark]

Oh, okay....

[NYMark]

A little Sephardic is possible but also seems very unlikely given the paper trail. I do have a brick wall at 2nd great-grandfather in Prussia (1st great was apparently illegitimate, and I have his mother's name but nothing further.) The Middle Eastern hasn't shown up on any other test....

[NYMark]

To return to the OP's question, it could simply be a flaw in MyOrigins that retesting would not correct. MyOrigins has me at 56% Ashkenazi and 3% Middle Eastern. I'm actually 50% AJ and almost certainly not more than that. Two close paternal relatives have tested with other companies and neither showed...

[NYMark]

For people with significant Ashkenazi ancestry, there's usually a relationship on multiple lines. I don't think any segment of less than 10 cM (and perhaps a good deal more than that) is worth paying attention to. I have a 29 cM segment match with someone that's clearly a mix of maternal and paternal...

[NYMark]

Some of these long blocks can be very persistent. This has been discussed on various threads over the years. The general consensus seems to be that FTDNA shouldn't count segments of under 5 cM. It leads to badly inflated predictions, especially when there's a significant amount of endogamy in a population....

[NYMark]

I would not count on this. The relationship can be a lot more remote, as far out as 8th or 9th cousin from what others have reported. In my own case, I can give two examples:

Longest block 22, total 55
Longest block 24, total 49

Both of these matches are related...

[NYMark]

I'd be really interested in seeing the size of second and third longest blocks for known relatives - first cousins and more distant – and also the totals with under 5 cM segments excluded.

[NYMark]

I'm half AJ. Half sibs. are 100% AJ and are also half-sibs to each other.

Me:
Half Sister: 109/1752
Half Brother: 101/1544

Half Sis to Half Bro:

72/1503

[NYMark]

I note that the OP said "a lot" of Ashkenazi matches. Of course, it's academic, as the ancestor is almost surely untraceable.

[NYMark]

A lot depends on the number of matches . . . but I still maintain that a remote AJ ancestor is the much likelier scenario, unless it's the number of matches is very small....

[NYMark]

Regarding the percentages, number of AJ matches is a way of validating whether it's noise or not. From following MiTuCents's thread on 23andMe, the best way to confirm it is to look at matches as well as the calculator. But I agree at very low levels, it's pretty speculative.

So the question...