Page 1082 of 1082 FirstFirst ... 8258298210321072108010811082
Results 10,811 to 10,814 of 10814

Thread: Could Western Jews (Ash. and Seph.) descend from Aegeans and Levantine admixture?

  1. #10811
    Registered Users
    Posts
    142
    Sex
    Ethnicity
    Venetian + CS Italian
    Y-DNA (P)
    I-Z58
    mtDNA (M)
    H7

    Republic of Venice Friuli Vatican
    Quote Originally Posted by Riverman View Post
    I think that's not enough because it especially swallows Slavic and Celtic. Celtic is somehow in between Germanic and Italic, so that's hard to model, especially for Italians I guess, but Slavic is a must, especially for the North East.
    Yep, I totally agree with you.

    Although the Northeast that I've used (thanks to Mokordo) is not just Friuli Venezia Giulia but also Trentino Sudtirol and Veneto.
    So for sure there is slavic admixture but also (and I believe mostly) germanic.
    And the slavic admixture probably didn't come as medieval pure slavic (which is basically the baltic-slavic that seabass used) but as a South Slav already admixed. Germanic influence also probably is not pure scandinavian-like but came admixed too.

    I agree with you though. It is tough, and currently impossible with DIY means at least, to distinguish among celtic, slavic, germanic ancestry.
    I think that the Seabass panel though is pretty well prepared as a starting point for fun (then once you analyze different populations you may want to play a bit with it to avoid weird situations as Greeks who have Iberian ancestry).

    We all dream, or at least I do, of a way to really estimate Hellenic, Italic, Germanic, Slavic, Celtic etc ancestry. But we'll probably need so many more samples and softwares...

    So for the time being, probably if we clean the panel to 4 populations the Germanic population must be read as something North of the Alps (which can be real Germanic, Celtic, Slavic etc... if the fit is not perfect you may guess that is a bit too western for some populations as North East or myself and you would need something a bit more Eastern).

    Still appreciated the effort and I think it does a good job with the available resources in showing how much Hellenic admixture is present in Italian modern populations (a bit overestimated I think, but still pretty good)
    Deep Ancestry: 63.2% AUT_LBK_N + 36.8% RUS_Vonyuchka_En @ 2.6339%
    Ancient: 89.0% ITA_Rome_Late_Antiquity + 11.0% HUN_Avar_Szolad @ 2.4649%

  2. The Following User Says Thank You to Northern Adriatic For This Useful Post:

     Cascio (05-06-2021)

  3. #10812
    Registered Users
    Posts
    2,781
    Sex
    Y-DNA (P)
    R1b-DF27 FGC17112
    mtDNA (M)
    H5a1

    Quote Originally Posted by Riverman View Post
    Ok, 1 percent, that's fairly low and can be explained either by problems of the algorithm, like passenger explained already, by minor Sephardic or other Jewish contributions, or simply by an unusual accumulation of that kind of ancestry. Very minor North African on 23andme being not that uncommon, even in Germans on the trace level usually. Its sometimes hard to tell whether its real or not. I would always try to trace back the segments in question to get a better impression of whether it could be real or not. Is it always the same segments, is it the same in the triangulated matches? That's what really matters.
    Some segments can change label depending on context, that's true even for Ashkenazi if its no safe segment. Then it can get assigned to Ashkenazi in a fully AJ person, English in an English, German in a German and Eastern European in a Pole. And you don't know for sure where it is from, if all those get a 100 percent of the regional ancestry. Yet its the same, usually smaller, segment. This just shows how the algorithm works with the smoothing effect, even for such rather better differentiated ancestries like AJ, NW and EE. Minor other contributions, which being not considered "safe segments" by the algorithm being swallowed and disappear in the majority ancestry. And its the right decision if the algorithm can't tell the assignment of a segment with any higher level of probability, let alone certainty.
    So that a 100 percent Ashkenazi person gets no very minor percentage is also no proof of anything, in my opinion, because AJ get smoothed out results as well.

    If the segment sticks at the higher and highest confidence levels and appears in other matches too, it might still not be real, but the algorithm definitely considered it real to put it that way. Usually, at least on 23andme, there is a reason for that decision. That might not tell you the ultimate source, but there is a connection between these two components in your relatives, whatever it may be, in which direction it might go. More likely is, if it sticks in relatives and at higher confidence, its non-Ashkenazi, but other Jewish in origin.
    Yes, I understand what you mean. I think their algorithm still needs work. For instance, I get overinflated French and German while a relative will have more of the Mediterranean mix. Maybe their Askenazi cluster is heavily skewed to region where less of our relatives were lost in WWII. Mine came from further west before the war.

  4. The Following User Says Thank You to Chad Rohlfsen For This Useful Post:

     Jatt1 (05-04-2021)

  5. #10813
    Registered Users
    Posts
    1,713
    Sex

    Quote Originally Posted by Chad Rohlfsen View Post
    Yes, I understand what you mean. I think their algorithm still needs work. For instance, I get overinflated French and German while a relative will have more of the Mediterranean mix. Maybe their Askenazi cluster is heavily skewed to region where less of our relatives were lost in WWII. Mine came from further west before the war.
    I also think, like I wrote some time ago, that there is a difference between the common founder segments, and more regional, community specific ones, which are only a rather smaller fraction and a good portion being detected because of the large reference for AJ, but still there are some of these below the radar, even for 23andme. So like with all other people, some subpopulations might be somewhat undersampled, even though that's much less a problem for AJ than for other ethnicities, because the reference is so big and because of the endogamy to the outside, but with constant gene flow to the inside, resulting in homogenisation and a solid assignment overall. My personal impression is that some regions and subpopulations had some gene flow which is not shared, both with other Jewish and non-Jewish populations and its these with which 23andme struggles sometimes. But its not necessarily a problem of the algorithm alone, but its these segments ancestral assignment per se, which is no easy task, because it being shared between different populations for many, many generations.
    For some of these cases, a more accurate assignment might only be achievable by comparing with ancient DNA, and even then you could ask: If the same ancestor and segment was present since before 1700 in different populations, and you claim you assign based on the recent ancestry, what are you supposed to do? It might be no recent admixture any more, its probably more than 300 years old. So telling someone he had this or that ancestry within the last 8 generations would be just wrong. Like people with a good tree might question the fidelity of ancestors and begin to search for NPE's and adoptions, just based on such a result. The timeline of 23andme just goes back to 1710, but there can be segments of significant size much older, at least back to about 1500. It might be not as common and not the rule, but its a reality. That's why trace results are always borderline, because even if they are real, they might not be recent but quite old.
    Last edited by Riverman; 05-04-2021 at 12:39 PM.

  6. The Following 2 Users Say Thank You to Riverman For This Useful Post:

     Cascio (05-06-2021),  Chad Rohlfsen (05-04-2021)

  7. #10814
    Registered Users
    Posts
    2,781
    Sex
    Y-DNA (P)
    R1b-DF27 FGC17112
    mtDNA (M)
    H5a1

    Yeah. It does have me wondering if my grandmother's grandmother, who is missing from records in Germany and the states, doesn't come from a different diaspora group.

    23andme also gave me an Algerian relative, but she doesn't know her dad, so I didn't pursue it any further.

Page 1082 of 1082 FirstFirst ... 8258298210321072108010811082

Similar Threads

  1. Replies: 20
    Last Post: 09-18-2020, 09:29 PM
  2. Do Mandaens descend from Jews?
    By StillWater in forum Jewish
    Replies: 1
    Last Post: 06-07-2020, 04:38 AM
  3. Late Gupta era admixture in Western India
    By tipirneni in forum Southern
    Replies: 85
    Last Post: 10-24-2019, 03:16 PM
  4. Replies: 3
    Last Post: 08-29-2019, 04:04 AM
  5. Replies: 18
    Last Post: 08-04-2019, 05:01 AM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •