No announcement yet.

How to tell genetic distance with older TMRCA

  • Filter
  • Time
  • Show
Clear All
new posts

  • How to tell genetic distance with older TMRCA

    I have what strongly appears to be a late medieval group of families. They have very similar haplotypes, and share the rare SNP Z17926, which is a subclade of a rare sister branch to DF29, in haplogroup I1. The haplogroup I1 project previously categorized them as the AS121210 haplotype cluster. They have a recognizable group of STR values that they share that set them apart, and Ken Nordvedt was who first recognized it. Not me. Now it and the I-A8100 project confusedly have one line of them with downstream SNPs, and another with a different downstream SNP, even though when I checked I found that they all in fact have all of those SNPs. I think maybe the project admins are confused about the endpoint SNPs because most people wouldn't submit their results to YFull.

    Ken Nordvedt thought the group's age is Saxon, but that's not what I"m seeing in the genetic distance, even by my own counting.

    There seems to be little interest or will among hte 50 families in the group to do the Big Y, though two lobes have tested and a third is in progress.

    What we appear to have is groupings of families, often with some degree of geographical clustering as well, who are clearly related to each other within four hundred or five hundred to seven hundred years, and several lone people with similar genetic distance, separated from each other by several degrees of genetic distance.

    There is no other clear pattern to the variation, such that, for instance, one could begin to build a hierarchical tree.

    One gets the idea that between 1100 and 1400 AD, somebody with the resources to do so were galloping around England, southern Scotland and Wales leaving babies behind. Someone who suddenly appeared in England between 1100 and 1400 AD came from northern France or medieval Flanders. All three Z17926 people who do not belong to this cluster come from the upper Rhine, central western France (specifically in an area populated by Celts who migrated from another location directly to the north), and eastern Scotland. The age of Z17926, and the directionality of I-A8100/ Z17954 is consistent with Celtic distribution; you'd expect to find this SNP in western France or medieval Flanders, if it wasn't already in eastern Britain, but nothing suggests that it was. Most tellingly, the people who WERE in Eastern Britain in 1100 AD didn't spend the next 300 years galloping around the island on horses, nor did they own the multiple manors that could have people traveling around and leaving babies all over the place.

    This family also had several geographical areas of common behavior. In Scotland there was no socioeconomic selection, unless that explains why the Earls of Hunter or whatever appear to share the DNA. The western Scottish group spread their Y DNA to coastal Wales. The East Anglian group are huge, dispersed, genetically similar - they may really ahve a gd of only 7 - and they contained no poor people, but gave rise to several successful Puritan families and wealthy and powerful Virginia gentry, who proceeded to contribute his own personal Y DNA to ten other very similar neighbors. One of the west Scottish group were extremely wealthy skinners of London in the mid 17th century; a member sat on the committee that ran Bermuda, and sent his brother to Bermuda, where he loaned his Y DNA to another family.

    I'm currently supposed to be helping the latest NPE of this group figure out how she connects to a Murray line that couldn't spell their name. She was able to identify her husband's father, but the line dead ends in Texas. "May you have interesting Y DNA".

    The problem I'm faced with, is Family Tree DNA keeps telling me that large chunks of this group, which exact people get picked up as matches varying from who in the group tested, have a genetic distance from each other of only 7.

    When I count, I get genetic distances across the group between 8 and 11, which is consistent with everything I just said about discrete subgroups and several SNPs separating them in time. It's also very consistent with suddenly appearing around or just after the 11th century.

    I've been over it with Family Tree DNA. Most of the difference in the genetic distance counting methods is specifically revolving around the behavior of cdy a and cdy b. Get people in this cluster with 9 to 11 points in genetic distance, and half of it is almost always changes in cdy a and cdy b. Occasionally there are also multiple changes or multiple point changes in DYS 464.

    Family Tree DNA counts changes in cdy a and cdy b as follows. For either cdy a or cdy b, multiple points of change are 1 point. Change in each of them counts as 1 point.

    So I guess, over thousands of years of genetic distance, Family Tree DNA is assuming cdy a and cdy b would have changed by a maximum of 2 points?

    I suspect that the truth is that Family Tree DNA's methodology doesn't take old family groups into account.

    My group is hardly alone; one of the Hamilton Y DNA groups is medieval in age, and its Norman origins are known. I've got a Doolittle family group that traces by paper trail to the 1400s, with a genetic distance across the group of 9. My brother in law's McKinstrys trace by paper trail to the 14th century, with genetic distance across the group of atleast 8. They all have the same haplotype but never all match each other within 7 degrees of genetic distance. The matches 85% overlap.

    I need to know how to accurately judge the TMRCA of this group.

    I'd appreciate ideas, and also referrals to who would have the expertise.


    Dora Smith

  • #2
    I recalled that a large Scottish Hamilton family is also medieval, and looked at its project page.

    If you look at group A, you will see a very similar pattern. There are maybe 175 people. There isn't much genetic variation. Most of the variation is in just three markers; cdy a and b and dys 570. There are also far more testees than make any sense with a recent TMRCA. The expected TMRCA is, according to the project administrators, 500 to 1000 years ago. The well documented ACTUAL TMRCA, is about 1270 AD. The identity of the founder of this group is well documented. He founded a powerful aristocratic family.

    Group B is almost as old and belongs to the same family, by a NPE that resulted in the founder's 2x great grandson. It has even less genetic variation and little variety in cdy.

    On the other hand, if you look at the McKinstry project that I administer, you'll see cdy go nuts when the genetic distance at 67 markers exceeds about 2, and the known TMRCA of this family is around 1400.

    Should we just figure if there's not a whole lot of genetic distance, cdy doubles the genetic distance to 10 all by itself, and reason leads one to believe the family group is more than 500 years old, it must be medieval?

    Or is there a more exact way to estimate it?

    I think they probably knew Hamilton group A was exactly 700 years old when they estimated its TMRCA.

    I've looked at some of the web sites and papers on genetic distance and TMRCA, and they're literally written in Greek. They also would lead one to believe that cdy undergoes one change every 28 generations, which is SO not true. Looking for common sense here.

    Thank you.


    • #3
      I left de Haas out of my equation. One of the haplotypes I discussed with Family Tree DNA was hers; her genetic distance from several of my project members is 6 at 67 markers, but I get between 9 and 11, and cdy accounts for half of the difference. She tested a male relative. She belongs to a Jewish family in the southeastern Netherlands. She has none of the much closer matches that most other people in this cluster have. This is consistent with the idea that her ancestors' last contact with an English source of Y DNA was in Plantagenet times. After about a century of increasingly serious mistreatment, Edward I expelled the Jews, who were most likely to go to the Netherlands and Germany, and then Edward III invaded France, which was disputed territory during the Hundred Years War. Before it was over, English armies had completely laid waste to much of Flanders and northern France. It makes sense to think her TMRCA is as Plantagenet as it looks. Also, while the Normans scattered Jewish communities across England so that they could provide coins to pay taxes with, they were concentrated where the Y DNA was - in East Anglia. There aren't supposed to have been more than 2000 of them.


      • #4
        Did you have any luck on fing more information on this SNP? I tested positive for it but have found very little information beyond your post.
        Thank you.