PDA

View Full Version : GEDmatch current matches



dp
11-17-2014, 03:53 PM
As anyone noticed a dip in the number of matches at a certain MRCA distance? and then a subsequent increase. I had assumed the number of matches should progressively increase as we look at matches with more distant MRCA.
Technique: Copy/Patse your GEDmatch match index to Excel and select on the lines for each gen (the count will be on the left in a little box). Then make an Excel chart of your results. One of those I chose was about 50th , the other one about 500th. :typing:
dp :-)
viz:
http://www.anthrogenica.com/attachment.php?attachmentid=3015&d=1416239210

Petr
11-17-2014, 06:28 PM
This is the result for my 14 persons:
http://i58.tinypic.com/2lv0ch.jpg

dp
11-17-2014, 06:33 PM
This is the result for my 14 persons:
http://i58.tinypic.com/2lv0ch.jpg
looks like 4.9-5.3 generations is a breakpoint in something for some reason.
Any ideas?
dp :-)
Your HK match has its maximum matches/gen a full generation sooner (more recent) than the other 12 you ran, and the 3 I ran. I guess it more closely matches the average GEDmatch kit in ancestry.
PS: It's nice being able to do direct comparisons like this B)

Petr
11-17-2014, 07:34 PM
This is other view to my GEDMATCH results:
http://i58.tinypic.com/2zjglyc.png
It seems that there is a clear correlation between Ashkenazi ancestry and the average distance. In Eurogenes Jtest and in 23andMe test these persons have the following percentage of Ashkenazi ancestry:
HK: 25,14 % - 63,5 %
BK: 16,17 % - 42,6 %
HB: 14,58 % - 34,6 %
JB: 9,04 % - 13,7 %
SM: 6,88 % - 13,2 %
others below 5 %.

I don't know what is the reason - maybe just Jews do more DNA testing than other populations.

In fact, I even don't believe that there are really so many so close relatives.

dp
11-17-2014, 07:54 PM
Maybe more endogamy in your graph's kits explains the generational shift.
I guess the dip (in our initial graphs) are effects of endogamy simulating/causing a genetic bottleneck in our graphs.
dp :-)

Salkin
11-17-2014, 08:52 PM
Well. I'm certainly getting a different result from you guys. This is from my FTDNA kit:

http://i.imgur.com/J7d8TYQ.png

And my 23andMe kit:

http://i.imgur.com/9SsuoJJ.png

Seven generations sounds a bit high to refer to the mass emigration events of the late-ish 19th century, but perhaps not impossible. I'm thinking a lot more descendants of those who left for the US have tested, compared to the descendants of those staying behind, maybe?

EDIT: Added 23andMe graph for comparison. It's interesting that the two differ noticeably. The amplitude of about everything is less for 23andMe, but then it does seem to get fewer matches.

dp
11-18-2014, 03:39 PM
Well. I'm certainly getting a different result from you guys. This is from my FTDNA kit:

http://i.imgur.com/J7d8TYQ.png

And my 23andMe kit:

http://i.imgur.com/9SsuoJJ.png

Seven generations sounds a bit high to refer to the mass emigration events of the late-ish 19th century, but perhaps not impossible. I'm thinking a lot more descendants of those who left for the US have tested, compared to the descendants of those staying behind, maybe?

EDIT: Added 23andMe graph for comparison. It's interesting that the two differ noticeably. The amplitude of about everything is less for 23andMe, but then it does seem to get fewer matches.

Yep, but you do have a peak at 5.2gens, which is basically where the kits we ran also have one. This mean you have a sudden dip after the point. It appears that there have not been many Swedish kits processed by either company yet.
dp :-)

dp
12-19-2014, 08:23 PM
I came across a kit with very few matches, well at least not 1500. In case your thinking it, based on the id number it doesnt appear to be a phased kit {starting with P} or one formed by Lazarus {starting with L}. With normal defaults (7cM) the kit had 204 matches. I lowered the minimum criteria. I found at that 5cM is the mininum that one-to-many will generate. This yielded a max GD=7.7 Even with that, the kit only had 680 matches.
I'm curious how few matches do some of Anthrogenica's users get, and what is the max GD? The kits I've tested with FamilyTreeDNA have a max GD of 7.1. A 23andme kit goes up to 7.4 generations. All have 1500 matches.
dp :typing:

Tolan
12-20-2014, 06:10 PM
I came across a kit with very few matches, well at least not 1500. In case your thinking it, based on the id number it doesnt appear to be a phased kit {starting with P} or one formed by Lazarus {starting with L}. With normal defaults (7cM) the kit had 204 matches. I lowered the minimum criteria. I found at that 5cM is the mininum that one-to-many will generate. This yielded a max GD=7.7 Even with that, the kit only had 680 matches.
I'm curious how few matches do some of Anthrogenica's users get, and what is the max GD? The kits I've tested with FamilyTreeDNA have a max GD of 7.1. A 23andme kit goes up to 7.4 generations. All have 1500 matches.
dp :typing:

I have 7.5 generation but it may be incomplete, I reach 1500 matches.
Here my graphic:
3258

dp
12-20-2014, 06:54 PM
I have 7.5 generation but it may be incomplete, I reach 1500 matches.
Here my graphic:
3258

I was surprised how flat the graph of your <7 gen matches were, especially considering your haplogroups, etc.
dp :-)

dp
12-22-2014, 08:02 PM
I came across a kit with very few matches, well at least not 1500. In case your thinking it, based on the id number it doesnt appear to be a phased kit {starting with P} or one formed by Lazarus {starting with L}. With normal defaults (7cM) the kit had 204 matches. I lowered the minimum criteria. I found at that 5cM is the mininum that one-to-many will generate. This yielded a max GD=7.7 Even with that, the kit only had 680 matches.
I'm curious how few matches do some of Anthrogenica's users get, and what is the max GD? The kits I've tested with FamilyTreeDNA have a max GD of 7.1. A 23andme kit goes up to 7.4 generations. All have 1500 matches.
dp :typing:

UPDATE: I ran HarappaWorld, Africa9 and EthioHelix K10 + French on the aforementioned kit. It is apparently of someone with 95% African ancestry. This means in a genealogical time frame the person only matches 200 people on GEDmatch. Apparently more kits are needed in GEDmatch from persons of African descent.
dp :-)
PS: I decided to post this admix info because their is a thread referring to African admix in another thread (http://www.anthrogenica.com/showthread.php?1796-quot-The-History-of-African-Gene-Flow-into-Southern-Europeans-Levantines-Jews-quot&p=25186&viewfull=1#post25186).