Having identified the probable ancestral origin of John Wilson (1716-1799) using Y-DNA in Part 1 of the blog post series, Part 2 uses autosomal DNA (at-DNA). The purpose here is to provide corroborating evidence that John’s ancestral origin is County Fermanagh in Northern Ireland as well as to offer greater geographic specificity within the county.
Compared to Y-DNA, at-DNA has the benefit of identifying more recent cousins (possibly as far back as six to eight generations). Also, both males and females can take the test enabling a larger pool of people who can ultimately help answer the research question.
For this analysis, I used a third cousin who is two generations closer than I am to John Wilson (1716-1799). John is her 4x great grandfather giving her potentially as much as 75% more DNA from earlier ancestors than I would have received. The objective of the at-DNA analysis is to find genetic clusters of DNA matches linked to the Wilson line and determine if the cluster(s) support the tentative County Fermanagh ancestral origin conclusion from the earlier Y-DNA analysis. Ancestry.com’s autosomal DNA product is used.
Genetic Cluster Identification
Two genetic clusters were identified using Ancestry.com’s “Shared Matches” feature using what I call an “earliest-known generation sibling search strategy”. To identify relevant clusters, this entailed viewing the shared matches for cousins who descend through other children of John Wilson (1716-1799) and his wife, Ann Skelton (1721-1803), other than the one my cousin and I descend. Unfortunately, all clusters here were associated with the Skelton side of the ancestral couple. Therefore, I moved down a generation through John and Ann (Skelton) Wilson’s son, William Wilson (d. 1804), through whom my cousin and I also descend.
This resulted in the identification of two viable clusters. Care was taken to ensure these clusters were not associated with the line of William Wilson’s wife, Phebe Penrose (d. 1842), but rather the Wilson line. This was done by reviewing the shared DNA matches for members of each cluster to ensure they were not associated with other matches who are confirmed to be part of our Penrose line. None appeared to be so.
As discussed shortly, I call the two identified clusters as the “Pennsylvania Cluster” and the “New Haven Cluster”, which are depicted below in a Gephi network graph (see Figure 4).1 The graph was created by downloading all my cousin’s in-common with matches and corresponding segment data from Ancestry.com using the DNAGedcom Client.2 From the resulting DNAGedcom file, only the members of the two identified Ancestry.com clusters were retained for analysis. DNA matches are sized according to the amount of shared DNA in centimorgans (cM).
The network graph delineates the two clusters. Highlighted are four matches who descend from children of William and Phebe Penrose Wilson other than the one my cousin descends, and these matches are labeled with either a 1 or 2. The Pennsylvania Cluster is a larger network, which is probably due to a couple of factors including the number of people who have tested and its members having immigrant ancestors who arrived the U.S. about two to three generations prior to those in the New Haven Cluster, which makes sense given that American Ancestry.com DNA test takers outnumber international test takers.
Prior to a brief introduction of the Pennsylvania and New Haven Clusters, Figure 5 theorizes how these clusters likely fit into John Wilson’s (d. 1799) extended County Fermanagh family tree. It serves as an additional reference point for understanding the many matches, sub-clusters, and documentary evidence yet to be presented. For a full discussion of the clusters, readers may review my final research report, which will be posted with the third and final blog post.3
Figure 6 graphically presents the DNA matches for the Pennsylvania Cluster. The cluster is comprised of 53 shared DNA matches forming five sub-clusters, which are organized using their family trees back to their own common ancestral pair rather than the theorized family tree presented in Figure 5. All sub-clusters match one another within the overall Pennsylvania Cluster. One sub-cluster (12 DNA matches) represents the DNA matches who descend from William Wilson (d. 1804) and Phebe Penrose (d. 1842). The other sub-clusters, which are named based on their ancestral origin of birth, include the so-called “County Fermanagh Cluster” (21 DNA matches), “Irish Cluster” (10 DNA matches), and a single match called “County Fermanagh Farry Match” (1 DNA match). The final sub-cluster (9 DNA matches, which is not graphically shown) consists of the “unknown match group” whose respective family trees are not publicly available and for whom I could not reconstruct a tree through genealogical means.
Across the three Irish sub-clusters (excluding the William Wilson Cluster), the amount of shared DNA varies from 44 cM to 8 cM with eight matches having more than 30 cM of shared DNA. According to Ancestry.com, these eight matches possess a high level of confidence (95%) of having a recent shared common ancestor with my cousin.4 The high confidence matches are found within each of the sub-clusters: County Fermanagh Cluster (5 high confidence DNA matches), County Fermanagh Farry Match (1 high confidence DNA match), and Irish Cluster (2 high confidence DNA matches).
The identification of the ancestral origins of the County Fermanagh Cluster comes from a biographical sketch of Robert J. Wilson (b. 1845), who is a great grandson of David Burgess Sr. The sketch states that Robert’s grandparents were:
“…John Wilson and Mary (Burgess) Wilson, natives of the Emerald Isle, but of Scotch descent. They were born in County Fermanagh, Ireland, emigrated to America shortly after the Revolutionary War, settling in Lancaster County, Pa., where they resided several years. Thence they removed to Mifflin County, [Pennsylvania] where most of the children were born and raised. Subsequently, they removed to Ohio…” 5
The County Fermanagh sub-cluster includes Mary Burgess, who married John Wilson.6 It also includes David Burgess Jr., who married a woman named Nancy.7 Because Nancy’s surname is unknown and Mary Burgess married a Wilson, I wondered whether Nancy might be a sister of John Wilson and whether the at-DNA connection is through the Burgess line and/or the Wilson line. As such, I engaged in targeted Y-DNA testing and recruited a male Wilson descendant of John and Mary (Burgess) Wilson and sponsored his Big Y-DNA test. Results confirm that he is a Y-DNA match to me, and the recruited Big Y test taker is graphically represented by “Wilson 10a” in Figure 2, which was presented in the earlier Part 1 blog post. While these results do not confirm that David Burgess Jr.’s wife Nancy is a Wilson, it does provide some support for the theorized County Fermanagh family tree for John Wilson (1716-1799) presented in Figure 5.
The Irish sub-cluster includes Henry Cosgrove who married Mary Jane Wilson.8 Both County Fermanagh and Irish sub-clusters lived about eight miles from one another in Mifflin County, Pennsylvania in the early 1800s and by 1830 all moved near Westchester, Ohio, which is town in Tuscarawas County, Ohio bordering Harrison County, Ohio. 9
For the County Fermanagh Farry Match, the paternal great grandmother of the test taker for this match emigrated from County Fermanagh to Philadelphia in the 1880s.10 Because this “sub-cluster” has only one match, it is difficult to ascertain for certain that the identified connection is through the Fermanagh great grandparent line. However, a thorough review of the test taker’s other lines finds no other possible connections to William Wilson (d. 1804) or Phebe Penrose. The great grandmother is Bridget Farry (1859-1921), who married Charles Carr in Philadelphia.11 Bridget Farry’s mother was Catherine Bannon, who resided near and died in Cavancarragh in County Fermanagh.12 The importance of the Bannon surname and Cavancarragh is discussed next and further supports the theorized County Fermanagh Wilson family tree presented in Figure 5.
New Haven Cluster
Figure 7 graphically presents the DNA matches for the New Haven Cluster. The cluster is comprised of 28 shared DNA matches forming three sub-clusters. One cluster (12 DNA matches) represents the DNA matches who descend from William Wilson (d. 1804) and Phebe Penrose. The other so-called “Cavancarragh, County Fermanagh Cluster” is comprised of five DNA matches. The remaining 11 matches constitutes “unknown matches” for whom trees were not available and/or their connection to the cluster could not be determined. Within the Cavancarragh sub-cluster, the amount of shared DNA varies from 37 cM to 11 cM with two matches having more than 30 cM of shared DNA. These two matches have a high level of confidence (95%) of having a recent shared common ancestor with the test taker.13 The common ancestral pair of the Cavancarragh Cluster is James Bannon III (b. 1829) and Bridget McHugh. The ancestral origins of James Bannon III’s wife, Bridget McHugh, is unknown.
Based on evidence presented in a family research report prepared by M.S. Pitkin,14 the Bannons appear to have immigrated to New Haven in a couple of waves during the 1840s and 1850s. James Bannon III (b. 1829), along with his parents, James Bannon II (b. 1800) and Margaret (last name unknown), and siblings, were enumerated in New Haven in 1850.15 James Bannon II’s (b. 1800) father, James Bannon I (b. 1767), appears to have immigrated in the 1850s and died 27 Mar 1858 in New Haven.16 James Bannon I (b. 1767), James Bannon II (b. 1800), and James Bannon I’s other son, Michael Bannon, appeared in the Tithe Applotment Books (agricultural tax records) for 1835 in Cavancarragh in Derryvullan Parish, County Fermanagh, which is within a couple of miles from several of the towns identified in the Big Y-DNA analysis presented in the earlier Part 1 blog post.17
Chromosome Mapping and Segment Triangulation
An important determination when analyzing shared DNA matches within a cluster is to ensure they share a portion of the same segment on a chromosome. Ancestry.com does not provide a chromosome mapping feature, and so I uploaded my cousin’s at-DNA to GEDmatch, MyHeritage, and FamilyTreeDNA, which provide this feature. While only a few of the shared Ancestry DNA matches were also on these sites, enough were present to identify matching segments. Triangulated segments were found for two contiguous segments on chromosome 7 as displayed in Figure 8. The shared DNA matches from the New Haven Cluster appear to be located at the lower end of chromosome 7, and DNA matches from the Pennsylvania Cluster are located immediately adjoining the New Haven Cluster. Situated within and across the lower and upper portions are descendants from William Wilson (d. 1804).
While reviewing the family trees for matches on MyHeritage and FamilyTreeDNA at the identified locations on chromosome 7, four new relevant matches were found. One is a new member of the County Fermanagh Cluster group, which is part of the larger Pennsylvania Cluster, descending from David Burgess Jr. (identified in Figure 8 as sharing 19.4 cM). The second new match also appears to fit within the Pennsylvania Cluster based on its location on chromosome 7 (31.6 cM match). Interestingly, this match has a direct paternal line back to a Patrick (Patt) Bannon who lived in Killee in County Fermanagh, which is an adjacent townland to Cavancarragh. According to this match’s family tree, Patrick was born about 1788 and died between 1852 and 1862.18
The other two matches have Wilson ancestry from County Fermanagh and are also situated at the same location on chromosome 7 as the Pennsylvania Cluster. One match (12.6 cM) descends from Mary (Wilson) Armstrong (1804-1866), a daughter of James Wilson (1769-1853) and Rebecca Coulter of Derryhillagh, County Fermanagh. The final new match (12.0 cM) is also located at the same location as the Pennsylvania Cluster and descends from Hall Price Wilson (1851-1924), a son of Alexander Wilson (1819-1905) and Rebecca Price of Lackaboy, County Fermanagh. Lackaboy is about one mile from Derryhillagh, which is where several Big Y-DNA matches have ancestral origins as presented earlier in Figure 2 in the Part 1 blog post.
Summary in Brief
Evidence from the at-DNA test appears to confirm County Fermanagh as the ancestral origin of John Wilson (1716-1799). In conjunction with the Y-DNA analysis, the at-DNA analysis additionally appears to preliminarily point to an area just to the east of Enniskillen in County Fermanagh near the townlands of Cavancarragh and Derryhillagh as a more specific location of John’s origin (see Figure 3 from the Part 1 blog post for a map of this area).
The next and final blog post utilizes documentary evidence and historical writings to corroborate prior DNA evidence and contextualize the results. Within this last post, a link to full my full research report will be made available.
- Gephi.org. (2021). Gephi version 0.9.2, available at www.gephi.org. Visualization settings: The layout or dispersion of nodes used ForceAtlas 2, and modularity class was used to partition and color the nodes into sub-clusters. To improve visualization, all DNA matches who descend from the test taker’s direct line from John Wilson (b. 1784) were removed as were DNA matches with fewer than six connections to another DNA match.
- DNAGedcom LLC. (2021). DNAGedcom Client, available at www.dnagedcom.com.
- Wilson, Rick T. (2022). Research Report on the Ancestral Origins of John Wilson, who died 1799 in Franconia, Montgomery County, Pennsylvania, USA, available at www.MyFamilyPattern.com.
- Ancestry.com (n.d.), What does the match confidence score mean? Retrieved 23 July 2021 from https://www.ancestry.com/cs/dna-help/matches/confidence.
- Chapman Brothers (1887), “Robert J. Wilson” in Portrait and Biographical Album of Linn County, Iowa, Chicago, IL: Chapman Brothers, p. 311-312.
- Tuscarawas County, Ohio, land deed, David and Nancy Burgess to Neal Morris (1829), Book 5, p. 478, Recorder of Deeds, New Philadelphia, Ohio.
- While the Henry Cosgrove’s marriage to Mary Jane Wilson is undocumented, Henry had a child named Wilson Cosgrove and a grandson named James Wilson Cosgrove: “Public Member Trees,” database, Ancestry.com (http://www.ancestry.com, accessed 2 April 2021), “Cosgrove” family tree by Tracy Cosgrove DeCarlo, profile for Henry Cosgrove (1768-1845).
- 1820 U.S. census, Mifflin County, Pennsylvania, population schedule, Union, p. 191, image 2 of 4, David Burgess; database with image, Ancestry (www.ancestry.com, accessed 12 September 2021); NARA microfilm publication M33, roll 102. And 1820 U.S. census, Tuscarawas County, Ohio, population schedule, Perry, p. 176, image 1 of 2, John Wilson and David Burgess; database with image, Ancestry (www.ancestry.com, accessed 12 September 2021); NARA microfilm publication M33, roll 95. And 1830 U.S. census, Tuscarawas County, Ohio, population schedule, Perry, p. 14-15, image 5-6 of 6, John Wilson and David Burgess; database with image, Ancestry (www.ancestry.com, accessed 12 September 2021); NARA microfilm publication M19, roll 141. And 1820 U.S. census, Mifflin County, Pennsylvania, population schedule, Armagh, p. 352, image 4 of 4, Henry Crossgrove; database with image, Ancestry (www.ancestry.com, accessed 12 September 2021); NARA microfilm publication M33, roll 102. And 1830 U.S. census, Harrison County, Ohio, population schedule, Freeport, p. 228, image 9 of 16, Henry Cosproue [Cosgrove]; database with image, Ancestry (www.ancestry.com, accessed 12 September 2021); Family History Library Film 337944.
- The Philadelphia Inquirer (1921, December 17), “Carr, Bridget E.”, p. 21, col. 7, Philadelphia, PA; online database, https://Newspapers.com, accessed 12 August 2021.
- Philadelphia, Pennsylvania, Marriage Index, 1885-1951, Bridget Ferry and Charles Carr (1887), license 14821; database, Ancestry (www.ancestry.com, accessed 12 September 2021); Clerk of Orphans Court, Philadelphia.
- Civil Records of Death, General Register Office Enniskillen, Catherine Farry (5 Jun 1879), Cavancarragh, No. 401, p. 63-1, database with image, Department of Tourism, Culture, Arts, Gaeltacht, Sport and Media (www.irishgenealogy.ie, accessed 1 January 2022).
- Ancestry.com, What does the match confidence score mean?
- Pitkin, Mary Stanford (2013), “The Descendants of James Bannon of Fermanagh, Northern Ireland,” unpublished family research report, p. 11-12; PDF in author’s files, p. 6.
- 1850 U.S. census, New Haven County, Connecticut, population schedule, New Haven, p. 219b, image 230 of 485, James Barnow [Bannon] in James Barnow [Bannon] household; database with image, Ancestry (www.ancestry.com, accessed 15 August 2021); NARA microfilm publication M432, roll 47.
- Pitkin (2013), “The Descendants of James Bannon of Fermanagh, Northern Ireland,” p. 11.
- “Ireland Tithe Applotment Books, 1823-1838,” Irish Genealogy Hub (http://www.irishgenealogyhub.com/fermanagh/tithe-applotment-books/parish-of-derryvullan.php, accessed 15 August 2021), > Fermanagh Genealogy > Derryvullan Tithe Applotments, James Bannon Sr., James Bannon Jr., and Michael Bannon, Cavancarragh, 1835.
- “Moore Web Site,” database, MyHeritage (www.myheritage.com, accessed 28 August 2021), person-150013_457807291, Patrick Bannon (1788-1852).