They match - but then they don't

[sweethome]

Hi everyone,

I have a group of four participants who all match at the 25-marker level. They've recently expanded their markers to 37 and now it appears that they are really two different groups. Two of the participants match 37/37 and the other two are only 30/37 to them (and they are 36/37 to each other).

Has anyone else had this happen? I've never grouped folks together if they matched only at 12 markers but I was pretty confident of a match at 23/25 or higher.

Thanks,
Priscilla Warren
Warren Surname Project

[Jim Honeychuck]

Hi everyone,

I have a group of four participants who all match at the 25-marker level. They've recently expanded their markers to 37 and now it appears that they are really two different groups. Two of the participants match 37/37 and the other two are only 30/37 to them (and they are 36/37 to each other).

Has anyone else had this happen? I've never grouped folks together if they matched only at 12 markers but I was pretty confident of a match at 23/25 or higher.

Thanks,
Priscilla Warren
Warren Surname Project

Are they R1b? I'm not researching R1b, but from reading various posts it is clear that even 37-marker matches which are R1b often fall apart when more markers are added.

Regards,
Jim

[sweethome]

Jim,

They are in fact, R1b. Specificially R1b1b2.

This is quite disconcerting. A lot of folks have made assumptions for their paper genealogy trail based on these results. Agh!!

Priscilla

[Larry Harvey]

They match

Hi Priscilla,
I experienced a similar situation when analyzing our Harvey cluster of 15 members all Hg I1-M253-AS6 all derived from three Scots-Irish Presbyterian brothers who came to colonial British Virginia circa 1700 as part of large wave of Ulster protestants seeking relief from the Church of England under Cromwell. They settled in rural Lunenburg Co VA circa 1740 on land granted to John Caldwell for his Cub Creek Presbyterian Church. He was also from Ulster and was of French Hugenout origin. I soon discovered a total of 6 associated families with 11 members (all Hg I1-M253-AS6) having nearly identical yDNA with matches up to 35/37. The largest was the Andersons (6) with singles from the Curtis, Bohon, Gritton, Watson, & Smith. Many of these families also showed up on the Lunenburg Co VA early tax records near the Cub Creek Presbyterian Church. The Andersons also had their roots in Ulster in County Londonderry and were Presyterians. This led me to believe this was a prime example of ethnic clustering along religious line in the British colony of Virginia.
Our yDNA seems to differ primarily with the DYS447 with the Harveys being '21' and theirs being '22'. Likewise, our DYS456 DYS607 DYS576 = 14 15 17 while theirs was DYS456 DYS607 DYS576 = 14 15 16. We all had the DYS464abcd of '14 14 15 15' and all had the DYS385ab of '13 15'.
Additionally, I ran the Cullen predictor for all our matches with the Harveys varying from 83% to 93% Hg I1-M253-AS6 while theirs varied from 79% to 88% of the same Hg. The Hg I1-M253-AS6 is most normally associated with the Danish Vikings which settled in the coastal areas of Ulster which later were part of the British trans-plantation efforts of Scots to Ulster.
This is a case of the paper geneaology & the yDNA placing these families in the same place with near identical yDNA. As such this creates more research lines to pursue and lends credence to making informed inferences. Good luck.

Larry Harvey
Harvey Surname Group
yDNA Hg I1-M253-AS6
mtDNA Hg I2
ySearch & mitoSearch: S3BHN

[Jim Barrett]

This is quite disconcerting. A lot of folks have made assumptions for their paper genealogy trail based on these results. Agh!! Priscilla

A jury doesn't come back with a verdict until they have heard all of the evidence being presented. Why would these people make assumptions on only part of the results?

With R1b1, it is very common for people to have 100's of 12 or 25 marker matches and have only a few by the time they get to 37 or 67. That is one of the reasons FTDNA offers the additional markers.

[gwenboucher]

They match

My husband is in a group of 6 known distant cousins. They match up to 25 markers but my husband varies from the other 2 tested for 36 markers by a genetic distance of 2. These 2 markers are displayed in red as explained by FTDNA:

"The markers in red have shown a faster mutation rate then the average, and therefore these markers are very helpful at splitting lineages into sub sets, or branches, within your family tree.

Explained another way, if you match exactly on all of the markers except for one or a few of the markers we have determined mutate more quickly, then despite the mutation this mismatch only slightly decreases the probability of two people in your surname group who match 11/12 or even 23/25 of not sharing a recent common ancestor."

[Dawn Ellis]

I never assign an R1b to a group on less than 37 markers...I have learned they easily fall apart between 25 and 37 markers.

[Robert Casey]

Just to add an important note on what happens when you upgrade to 67 markers. This upgrade will really bring your two clusters together or separates them even more. For one Casey cluster, 12 submissions can in with the exact marker values from 38 to 67. These submissions are different lines where no traditional connections can be made and their oldest proven ancestor were born between 1770 and 1830 (average 8 or 9 generations). The MRCA calculator obviously revealed these line were much more closely related at 67 markers than at 37 markers. This cluster had four mutations from the MRCA for markers 1 to 25 (implied very closely related - only 3 submissions had any mutations). It has ten mutations for markers 26 to 37 (put some distance between them and revealed half belonged to one branch and the other half belonged to another branch).

For another cluster where the markers from 1 to 37 were less closely related, there were two mutations found in markers 38 to 67 (their closeness in relationship via MRCA calculator did not change much). There are only six submissions in this cluster (two were Genealogy.com that are missing all but two of the 38 to 67 markers).

DNA is a game of probabilities - the larger the sample size - the more accurate the results (both in number of markers and number of submissions). 12 markers is like building your family tree based only with tax records. 25 markers you get access only census records, 37 markers throw in all other legal documents available. 67 is like looking for very specialized records - sometimes a gold mine (finding new branches) - other times not a lot of progress (no or few mutations which are probably recent mutations vs. ancestor significant mutations).