If you’re using genetic networks to analyze your DNA matches, then you need to also use viewed match switching. This DNA analysis strategy helps you find more relevant matches for the ancestor being researched and construct a more robust proof argument.
I first introduced viewed match switching in my genetic network blog series in its second post, which explained how to identify misclassified matches. While this post is not about misclassification, it is about genetic networks. So, as a quick review, genetic networks are a group of people who share DNA, and these matches are visualized in the graphic below.
I label genetic network members as “A” (the test taker), “B” (the match being viewed), and “C” (matches that match both A and B). As described in an earlier blog post, you use strategic tree triangulation to select a “B” match that represents an ancestor or ancestral couple being researched. The “B” match should only share the ancestor(s) of interest with the test taker (A) and share no other common ancestor.
Ideally, the “B” match should also descend through a different child line of the ancestral couple you’re investigating than the child line that the DNA tester (A) descends (see the yellow ancestral lines in the image below, which assumes that the pinkish ancestral couple at generation five is unknown and for whom the research is attempting to discover). Finding a “B” match descending from another child line isolates the shared matches (C) to the ancestral line being researched (yellow couple at generation four). As such, these shared matches (C1, C2, and C3 below) only share ancestry that both “A” and “B” share together, i.e., great grandparents at generation four. Especially when “C” matches are ancestral cousins (i.e., C3), shared matches help take the ancestral line back another generation or more (see the pinkish ancestors at generation five and the cousin types in the image below or as described in an earlier post).
Viewed Match Switching
After creating the initial genetic network, the value of the network can be enhanced with a slight, albeit strategic, alteration. Viewed match switching involves swapping out the original “B” match that created the network with one of the “C” matches. Doing so enables you to expand the genetic network to include other cousins who match the tester (A) and some of the other “C” matches in the initial network. The newly expanded network should include additional “C” matches providing greater insight into the ancestor(s) of interest represented by the initial A-B-C genetic network.
Explained another way, viewed match switching expands the genetic network similar to the appearance that a Venn diagram portrays in the image below. In this example, the initial genetic network for the DNA tester (A) is in green with the viewed match (B) at its center and the shared matches between “A” and “B” represented by C1 through C10. Reviewing and building out the family trees for the “C” matches helps to determine how they are genetically related to “A” and “B”.
Before explaining how viewed match switching is performed, an additional concept needs to be introduced: unlinked family clusters.
Unlinked Family Clusters
Frequently, several of the shared matches form an unlinked family cluster, which is a relatively large group of “C” matches who all descend from a single ancestor or ancestral couple but to whom you are unable to establish your genetic relationship. To be a proper unlinked family cluster, the subgroup of “C” matches should descend through multiple child lines of their common ancestor within the cluster. Unlinked family clusters often descend from a previously unknown descendant cousin or one of your investigated ancestors’ parents or grandparents several generations back. Finding the connection between the DNA tester (A) and unlinked family cluster is how you break through brick walls.
Using Viewed Match Switching
Now, to expand the genetic network using viewed match switching, one of the initial “C” matches is selected within an unlinked family cluster such that it represents the largest match as measured by shared centiMorgans (cM). The selected “C” match then becomes the new viewed match (B), which is labeled as B prime (B’) in the previously presented diagram.
In the above example, C9 is selected as the new viewed match, and clicking the shared match filter within a DNA testing website visualizes the shared matches between “A” and C9 (now B’). The new network is in orange. Note that C8, C9, and C10 are members of both the original green genetic network and the new orange network. You want some members of the original green genetic network to be part of the new orange network as this is one indicator that these new matches are valid members of the larger genetic network.
Viewed match switching is similar in concept to Gephi network graphs, which takes considerable effort and time to create. Network graphs visualize your shared matches in their totality and in relationship to one another, whereas viewed match switching analyzes a portion of an ancestor’s genetic network one set of related networks at a time. For someone who does not wish or cannot create Gephi network graphs, viewed match switching is a great alternative using only the shared match filter within DNA testing website and strategic selection of a viewed match (B).
To better understand the value of viewed match switching, I put it into context with an example from my own research.
Viewed Match Switching Example: Hill Genetic Network
In Part 5 and Part 7 of the genetic network blog series, I introduced a genetic network for my 4x great grandfather, William Hill (1775-1836). The original network contained 105 shared DNA matches (C) from which I was able to identify four unlinked family clusters represented by the surname of their respective common ancestor, i.e., Keel, Clark, Linn, and Harris. Researching these clusters pointed toward a Hill family from Dorchester County, Maryland who migrated to Kent County, Delaware and then into Pennsylvania (see Part 7 for how this determination was made).
Using viewed match switching for this blog post example, I selected the largest match (44 cM) in the original network, which was a member of the Keel unlinked family cluster. This Keel match became the new “B” match (i.e., B’). After selecting the shared match filter, the new genetic network between and A and B’ contained 107 shared matches, of which 64% were members of the original network (i.e., Hill, Keel, Clark, Linn, and Harris) providing assurance that the new network was an expanded network and not an entirely new and unrelated network.
Remarkably, 39 matches were new providing an expanded opportunity to learn more about my research subject. After reviewing and building out the family trees for the new matches and conducting preliminary documentary research, I discovered two new unlinked family clusters with the common ancestral couple for each noted below:
- Thomas Vickers (1851-1930) and Margaret Cox (1857-1910) through three different child lines (7 total matches).
- James Matkins (1794-1862) and Rhoda Murray (1795-1867) through two different child lines (3 total matches).
While I cannot yet discern the connection to the Vickers/Cox couple, the surname Matkins is familiar to me and a rarer surname as well. Research on my part identified that James Matkins (1794-1862)[1] reportedly descends from John Matkins (1762-1843), who served in the Revolutionary War from Dorchester County, Maryland[2], which is where the initial genetic network analysis for my Hill line in the the earlier blog post pointed to geographically. Also of note is that one of the child lines represented by James Matkins in the new network is through a son named John Hill Matkins[3] (1817-1887) – emphasis on the middle name of “Hill”.
While the parents of John Matkins (1762-1843) are unknown, it is probable he descends through one of two early 1700s marriages between the Matkins and the Hills in Dorchester County. These Hills are the same family to which my 4x great grandfather William Hill is thought to also descend based on my earlier research. Hannah Hill (b. 1679) married Theodore Matkins,[4] and Hannah’s niece Sarah Hill married an unknown Matkins.[5] Coupled with the Matkins unlinked family cluster and one of its descendants named John Hill Matkins, there’s now fairly good DNA evidence that my William Hill (1770-1836) also likely descends from the Hills of Dorchester County, Maryland, who initially arrived Dorchester about 1669.
Conclusion
When you find an initial genetic network that looks and/or is promising, don’t stop there. Use viewed match switching to expand the network to find other matches providing even greater insight into your research question. Especially when you are working on building out your tree for ancestors further back in time, viewed match switching may help you add additional indirect DNA evidence providing you with reasonably exhaustive research to confirm a theory. Researching the unlinked family clusters found within the genetic network may even help you find direct documentary evidence, too.
The next blog post, which will be posted in several weeks time, will present and analyze my Hill genetic network using Gephi network graphs as a complement to viewed match switching.
Sources
[1] “Public Member Trees,” database, Ancestry.com (http://www.ancestry.com, accessed 18 December 2024), “Pearson Family Tree” by Meredith Trible, profile for James Matkins (1794-1862).
[2] Pension Application, John Matkins, Private, Revolutionary War, “Declaration of John Matkins in order to obtain the benefit of the Act of Congress passed June 7,1832”, dated 6 November 1832, Pension Application S.7181, Pension Office, War Department, Washington, DC; online database with images, Fold3 (www.fold3.com, accessed 18 December 2024).
[3] North Carolina, U.S., Death Certificates, 1909-1976, Mrs. Jeannette Isabel Matkins Jones (1957), certificate 14012, Alamance County (June 1957), image 13 of 47; database with image, Ancestry (www.ancestry.com, accessed 18 December 2024); citing North Carolina Archives, microfilm S.123, roll 19-242.
[4] Maryland, U.S., Wills and Probate Records, 1635-1777, Amy Hill (1739), Dorchester County, Vol. 22 (Book DD1), p. 125-126, image 609-610 of 823; database with image, Ancestry (www.ancestry.com, accessed 18 December 2024); citing Maryland County, District, and Probate Courts.
[5] Dorchester County, Maryland, will, William Hill (1733), volume 20, p. 884-885, Prerogative Court, Cambridge; online database with images, Maryland State Archives (www.msa.maryland.gov, accessed 18 December 2024), will series S538-30, book CC3 (1732-1734), image 464-465 of 518.