Judith

12-10-2016, 11:57 AM

There are some suggestions on this web site that the X chromosome can be particularly significant in finding recent (<300 years ago) ancestors. I have not found it so despite following the method suggested, my X matches had no autosomal DNA match and no uploaded GEDcom to help. But it may be more illuminating in other people trees.

Like the theory I share 196cM with my mother, and on the purely maternal line the X chromosome splits into approximately two on each generation. So after 7 more generations this will be down to just over 1.5cM (using a divide by 128). The line which divides least on the X is the male female male female, which needs 16 generations to divide to 1.5cM. And in 8 generations on the male female, male female it is down to 12cM. So is not possible to use X matches to predict the number of generations ago.

Like the autosome the X does not divide evenly and this can be illustrated nicely using ancient DNA. I chose BR2 (GEDmatch F999933) since we already know that she has autosomal matches with living people. The result of a one to many search is 32 matches. The table did not post well but it is easy to duplicate if you wished. Decreasing the threshold from 5 to 3cM does not give any more matches, it must be internally limited on the site.

The largest match is 10.5cM the smallest is 5cM. All but one of the matches are female. More than half have autosomal matches which the algorithm predicts to be 7 generations. But she is not! At 3200years old she is 128 generations ago. If you assume an exact 50% DNA split this 10.5cM could be from 5 to 8 generations ago.

So I have concluded that although X matches show where a match is in your tree, they are not otherwise useful.

Like the theory I share 196cM with my mother, and on the purely maternal line the X chromosome splits into approximately two on each generation. So after 7 more generations this will be down to just over 1.5cM (using a divide by 128). The line which divides least on the X is the male female male female, which needs 16 generations to divide to 1.5cM. And in 8 generations on the male female, male female it is down to 12cM. So is not possible to use X matches to predict the number of generations ago.

Like the autosome the X does not divide evenly and this can be illustrated nicely using ancient DNA. I chose BR2 (GEDmatch F999933) since we already know that she has autosomal matches with living people. The result of a one to many search is 32 matches. The table did not post well but it is easy to duplicate if you wished. Decreasing the threshold from 5 to 3cM does not give any more matches, it must be internally limited on the site.

The largest match is 10.5cM the smallest is 5cM. All but one of the matches are female. More than half have autosomal matches which the algorithm predicts to be 7 generations. But she is not! At 3200years old she is 128 generations ago. If you assume an exact 50% DNA split this 10.5cM could be from 5 to 8 generations ago.

So I have concluded that although X matches show where a match is in your tree, they are not otherwise useful.