Back in 2012, the original cluster analysis was completed using Genetic Affairs website. A lot of change has happened since that time so it seems appropriate to reviews these unknown matches to see if anything more can be learn from them.
- Genetic Affairs can no longer directly generate cluster diagrams due to a dispute over rights. Too bad, since
- Ancestry added a feature where you can see the DNA sorted by parents. In WAB’s case, parent 1 would be Baker/Woods while parent 2 represents Perkins/Nicklos. It’s not 100% accurate, but it is generally useful.
- Additional family has tested. WAB sister, and two first cousins were kind enough to test. In addition I have access of a second cousin
- Recently, ancestryDNA enhanced their shared DNA matches so that see the match from the a different viewpoint.
With all these new people and tools, I thought I would reviews that original Genetic affairs cluster diagrams, and see if we can solve any mysteries
In Part I, clusters for Woods, Baker and Perkins were isolated. In Part 2, I attempted to identify unknown clusters 20, 21 and 22 along with some unclustered single matches. Cluster 20 was lumped with the Perkins subgroup due to one gray dot shared by cluster 15 and Cluster 20. Cluster 21 was unidentified. and cluster 22 was correctly identified as the Nicklos cluster.
Cluster 21:
Let’s start with cluster 21 which turns out to be pretty easy. WAB shares matches with BM, CS-AM, BP761 and N&N in this cluster. The shared match list is shown below for BM. BM shares 1476 with CS-AM, 232 with GP761 and 29 cM with N&N.
If you look up the shared matches of WAB with CS-AM, GP71 and N&N, you can fill out the relationship values in the cluster.
Key to Ancestry’s new shared match list
- Amount of DNA shared by WAB and CS-AM along with predicted relationship
- Is this match predicted to relate to Parent 1 or Parent 2. In this case parent 1 is Baker/Woods.
- Amount of DNA shared by BM and CS-AM. In this case, the two individuals are closely related
- Indicates whether this person provided a tree. Trees vary widely. They may or may not be useful
- Here ancestry indicates that it has a predicted common ancestor that shows up in Thrulines
- The old limit for shared matches for WAB was 20 cM. Now ancestry shows matches down to 8 cM as long as BM also has a match at 20 cM or above.
It turns out that RS gives the answer to the puzzle. WAB and RS have a common ancestor of Isaac Hill and Ruth Squires, part of the Baker Line. Now can I give additional evidence with the DNA using any useful trees of the other matches.
I was able to turn BM from red to green solely based on their user name (Mellor), and the posted obituary of Doris Eleanor (Powers) Mellor.
Notice that I pick up two new matches: BH-BL and RS which lie under the old 20 cM limits for shared matches. I have added them to the cluster.
Cluster 20:
Ancestry indicates that all these matches are related to Parent 1 suggesting the cluster above is either a Baker or a Woods. The results conflicts with my original analysis showing this cluster part of the Perkins subgroup. The following drawings shows the steps in preparing the cluster for analysis
Purple Circles
CH and SJ are closely related and each have complete trees. Their common ancestors are Thomas R Johnston (1883-1967) and Kathreen R Grosvenor (1894-1980). The parents of Thomas R Johnston were born in Ireland. Kathreen’s mother was born England, while Kathleen’s father came out of New England.
Red Circles
For example, a search of HC-AIU matches shows RK who shares 2600 cM making them likely sister. Yet RK only shares 10 cM with WAB. Furthermore, RK has a useful tree we can examine.
HA-AIU, MCONV and RK all likely sisters. One of them has a published tree. All tree share around 100 cM of DNA with VM1. The common ancestor is William Magee (1827-1909) b. Ireland and Martha Moore (1830-1900) b. Ireland.
Blue Circles
EP, TR, TU-MLU, LA, JDLIZ82, MB and BR all cross the 50 cM level making them relatively closely related.
TU-MLU, JDLIZ82 and PG all share over 1000 cM.
LA and BR share over a 1000 cM.
BR & LA share 1687 cM. Common ancestors Gene Arlie Ayers (1938-2002) b. TN & Leona Stevens (1940-1996) b. MI
MH & EP share 41 cM. Common ancestor LaVern Darrel Calvin (1877-1958) b. PA and Bertha Orlean McCurdy (1878-1915) b. PA
The Irish connection of the purple and red groups suggests that this is a Woods cluster. The Woods family ancestors arrived from Ireland. around 1800 and settled Pennsylvania. Names in include Woods, McLaughlin, McDowell, McCutheon. Since Ireland represents such a difficult hurtle, it would be nice to see Pennsylvania connection in the blue group.
The Irish connection of the purple and red groups suggests that this is a Woods cluster. The Woods family ancestors arrived from Ireland. around 1800 and settled Pennsylvania. Names in include Woods, McLaughlin, McDowell, McCutheon. Since Ireland represents such a difficult hurtle, it would be nice to see Pennsylvania connection in the blue group.
Single Match: ScotterMac
It is possible to create a cluster from a single match, and apply the same technique. Here are results when I took the shared match list of unidentifed user ScotterMac (abbreviated SM) and turned it into its own cluster. The resulting cluster is not well formed, but it does allow for analysis.
All of these matches belong to parent 2 (Nicklos/Perkins). If Perkins, we would expect to see Canadian moving back to Scotland, New England, Pennsylvania or possibly Irish. If Nicklos, we would anticipate Saxony in Germany and Scotland. Using the same technique as outline above, three different DNA groups were identified
- Purple Circles shared by MR and KM39497. Neither has a tree so we have nowhere to go.
- Blue Circles shared by SA, DC-A and A4D6G. Common ancestors Robert A Grabau (1912-1986) of b. NY & Doris R. Hunter (1916-1966) b. NY
- Red Circles shared by SM, ES and CB. CB has a nice tree where you can see relative from Canada. SM also appears to come from Canada but the connection to SM has not yet been uncovered.
Single Match: WW-PB (37 cM over 3 segments)
This single match remains unknown. To create a table similar to above would require over 260 rows and columns. Here is a distribution of the shared match list between WAB and WW-PB. Of the 261 ancestryDNA assigned 250 to parent 1, 9 to parent 2 and 2 were unassigned. So this match is likely Baker/Woods.
| Shared Match Value | Count |
|---|---|
| 20 cM to 29 cM | 184 |
| 30 cM to 39 cM | 55 |
| 40 cM to 49 cM | 14 |
| 50 cM to 99 cM | 5 |
| 100 cM to 499 cM | 3 |
| 500 cm and up | 0 |
| Total Shared Matches | 261 |