Using DNA Painter in a Descent Study




In 1978, I began a genealogical journey that concentrated on the progeny of William Owston (1778-1857) and his wife, Frances Wilson (1782-1853). At the time, I was tracing the descent of four individuals with the Owston surname: Thomas (1804-1874), James Wilson (1809-1858), Euphemia (1824-1892), and John Gillon (1826-1901). The three men settled in the Greater Pittsburgh area and Euphemia in Seneca County, Ohio.  Following her death, her widower, their son and daughter-in-law, and two grandchildren moved to Pittsburgh.

By the end of 1978, I was able to determine that Thomas and James were brothers and Euphemia and John were siblings, but I was unable to determine their connection to each other and to an additional individual named John Conrad Owston (1851-1923) who also appeared in the records of Allegheny County, Pennsylvania.

In 1988, my research concluded that Thomas, James, Euphemia, and John were siblings and were the children of William and Frances Owston who settled in Upper Canada (Ontario) in 1820. William, a Royal Navy master, and his eldest son Thomas, a midshipman, qualified for land in the New World – specifically in Asphodel and Smith Townships in what is now Peterborough County.  After they sold their original holdings, the family moved to Hamilton Township in Northumberland County near the town of Cobourg in the early 1830s.


William and Frances Owston painted between 1848 and 1852 possibly by Jane Gillon Owston (1831-1921), their oldest grandchild to survive childhood.

In assessing local Canadian records, four additional siblings emerged over the next year: William Owston, Jr. (1807-1892); Frances Janet Owston (1815-1902); Charles Paget Herbert Owston (1817-1858); and Mary Ann Margaret Owston (1819-1909). Photographic evidence discovered in 1990 suggested that Charles P.H. Owston was the father of John Conrad Owston. Other evidence indicated that Charles had been in the Pittsburgh area as early as 1838.

Currently, I have identified 585 descendants of William and Frances (in all generations). There are obviously more, but I have not yet discovered and cataloged these individuals. At that, the number of William and Frances’s progeny is very, very small by comparison to other families at this depth. Much of this is due to an absence of issue in subsequent generations – a problem that is typical in all Owston/Ouston families.


In 2010, I began requesting descendants of William and Frances Owston to test. I concentrated on testing fourth cousins; however, other generations are present in the study. We currently have 36 tested members of the Cobourg line of Owstons. All 36 have tested autosomally. Additionally, four tested with the Big-Y500 and three pursued STR testing. Three have ancestral non-paternity events. Two are deceased (C-H & C-I).

Cobourg line participants of the Owston/Ouston DNA project descend from six of William and Frances’s children. Missing are the descendants of Mary Ann and Euphemia.  Two of Mary Ann’s children, William and Henry H. Smith and, their double first cousin – one of Euphemia’s sons, Henry B. Smith have been impossible to locate. The remaining children from these two daughters have no living descendants. While the lineages through the two sisters may be extinct, I cannot affirm that conclusion at the present and am hesitant to do so.

Aliases are used for all individuals and named by using Cobourg as a family identifier and an alphanumeric designation. Surnamed males are used with numbers, such as Cobourg11 (or C11) to comply with the identification of our Y-DNA project. These range from C01 to C13.

Females and non-surnamed males are identified by a hyphen and a letter, as in Cobourg-L or C-L. Currently, there are 20 individuals that fit this category (C-A to C-T).  Additionally, we have three individuals who cannot be compared with all participating members, as they have not uploaded their data to GEDMatch. They only can be compared with those on the same testing platform. These are listed with an X and a number, as with CX2.

There are five additional individuals who have tested with Ancestry and have not uploaded to GEDMatch; their segment data cannot be compared at the present. Four of these descend from John Gillon Owston and one from William Owston, Jr. They are not included in the project at the present.

The compared individuals are identified in the chart below:


In every descent line, only one grandchild of William and Frances is currently represented. In regard to William, Jr.; James Wilson; and Charles P.H., they have no other children from which there remains living descendants. Both Thomas and John G. have an additional genetic child that has living issue and Frances J. has three additional children who have living descendants.  None of these have been secured for testing at the present.

The family of the youngest child, John Gillon Owston, is overly represented. This is the author’s family that includes 16 additional individuals, 14 of whom are children and grandchildren of C01, C02, C03, and C-R. The remaining two individuals are descended from John G. Owston’s adopted son.


A while back I decided to experiment with Johnny Perl’s wonderful DNA Painter tool. In using it to paint my own DNA matches with relatives, I realized the possibility of this tool to chart the descent of couples and I set out to do just that. Since I had been contemplating updating the analysis of fourth cousins and other near distant relatives that I had conducted in 2015, this seemed like the perfect opportunity with this excellent tool. Seventeen additional Cobourg line participants have been added to the project since August 2015; however, only six are germane to this specific analysis.

I originally charted the matching segments of those who represented the oldest participant in his or her direct line, as our project has several children and grandchildren who have tested. This narrowed the field down to 18 individuals – eliminating the three individuals who had NPEs and 15 others who were downstream of older generation participants.

The majority of those charted are five generations from our common ancestors and are fourth cousins interlines. One, Cobourg-J, was four generations from William and Frances. Four were six generations out: C-D, C-T, CX1, and CX3. Eventually, I added Cobourg-A, the daughter of Cobourg-J, as she was a fourth cousin to many of the other participants, although her mother carried the same along with additional matching markers. Cobourg-A was also our fourth participant in 2010.


No relationships that were closer than third cousin, once removed were charted – as these would have been descended from the same child of William and Frances Owston.  The matches could be from either William or Frances or be combined segments of both ancestors. One hundred and fifty-three segments have been painted and, if this had represented the genome of one person, it would constitute 12%.

When adding the matching segments, I did the following:

I allowed matches at 5 cM and above, as my previous study on fourth cousins indicated that segments of that length were plenteous at the fourth cousin level. I realize that this move is a bit controversial, but sometimes the smaller segments triangulated with other matches that had larger segments which could be traced to William and Frances.


In naming the DNA match, I used the two people with whom I was comparing.

For the ancestor’s name, I used the names of each of William and Frances’s two children that were being compared.

To save space, I used “Paternal” for the side of the family of the match.  By not specifying either paternal or maternal, the matching segments appear vertically larger.

I picked a unique color for the descent of two children being compared.  I originally tried to have a unique color for each pair of individuals being compared, but that was laborious, and I reduced the total colors to the manageable number of 13. Looking at the individual chromosomes, you can compare the different combinations by color.  This makes it easier to find triangulated segments by looking for three different colors.

Triangulated Segments

Example 1:

The following example from Chromosome 3 indicates an easily identified triangulated segment from three different siblings:


The triangulated portion of approximately 17.7 cM corresponds to the pink segment and is the largest triangulated segment for the study.

In addition, there are two other smaller non-triangulated matches in blue between a descendant of Thomas (C-P) and a mother and daughter pair (C-J & C-A) through William, Jr.

Example 2:

The second largest triangulated segment, at approximately 14.3 cM, is between two second cousins who are second great-grandchildren of John G. (C-H & C-R) with descendants from Thomas (C-E) and Charles P.H. (C-G).  This example is from Chromosome 8.


Example 3:

On Chromosome 2, there is an approximate 8.9 cM triangulated match between descendants of Frances J. (CX3), Charles P.H. (C05 & C-G), and John G. (C02).


Examples 4 & 5:

Two triangulated segments of approximately 8.5 cM each appear at the end of Chromosome 10.  The first segment is shared by descendants of Thomas (C-E), William (C-T), and James W. (C08). The second is triangulated among the progeny of William (C-J), James W. (C08), and John G. (C01).

You’ll notice an additional match between C01 and U3 painted in mint green. This will be discussed later.


Examples 6 & 7:

Chromosome 7 has more matching segments than any other – a total of 35 with 10 different combinations representing the six children of William and Frances. These segments range from 6.1 to 32.7 cM.  Very early in this study, I realized that Chromosome 7 was important, as it kept showing up in various matches.

The end of the Chromosome 7 has two triangulated segments. The first shows a shared segment between descendants of Thomas, James W., and John G. at 6.5 cM. Missing is a match between CX1 and C08. This is because CX1 has not yet uploaded her data to GEDMatch and her 23andMe results cannot be compared with C08’s FTDNA test.


At the very end of Chromosome 7, we have a small segment that triangulates descendants of four children of William and Frances Owston:  Thomas (CX1 & C-Q), William (C-A & C-J), Charles P.H. (C05), and John G. (C03).  Note: CX1 is the niece of C-Q and C-A is the daughter of C-J.

Using the oldest generation represented from each son of William and Frances, the triangulated segment is at 7.7 cM.  Eliminating the smallest matches with Cobourg05 (Charles P.H.), the triangulated matches with the three remaining brothers appears to be approximately 12.7 cM.


While there are seven segments that can be triangulated among three and four of William and Frances Owston’s children, most matching segments are between descendants of two of the children or one with two individuals descended from the same child. These are indicative of the genetic network of the participants and their families.

This is illustrated in the segments found at the beginning of Chromosome 7.  There are multiple participants in single families that share segments with one of the other families.  C01, C02, and C03 are brothers and are second cousins to C06 and C-R. C06 and C-R are first cousins. C-I and C-Q are second cousins.


This example suggests the recombination points that occurred in earlier generations; some of these segments probably originated with William and some with Frances.  This is evident with the long mauve segment and the four brown segments.  There is no triangulation between the mauve and brown segments; therefore, a conclusion might be rendered that the mauve segment originated with a different ancestral parent than the brown segments.

An interesting phenomenon is found on Chromosome 11 where all the matches are green and show the connection between descendants of William and John G.  As previously noted, C01 and C02 are brothers and C-R is their second cousin. In William’s family, C-J is the mother and C-A the daughter while C-T is their respective half first cousin, twice removed and half second cousin, once removed.


The matching segments between C-T and the two Owston brothers (C01 and C02) are longer than their matches with C-A (a generation closer) and her mother C-J (two generations closer).  In the chart below, several of John G.’s descendants have larger matches with those in William’s family.


This begs the question, “are there other connections between these two families or any of two families in the study?”  Between William and John’s descendants, there are not. Between other descendants of William and Frances, there are additional shared ancestries; however, none of these are represented in this study at the present. As more people test, it is likely these other shared ancestries will emerge and the total amount of cM will enlarge with these participants.


During the past year, we have encountered a family that matches the Cobourg Owstons, and more specifically the descendants of John Gillon Owston.  Two individuals who have tested with Ancestry match brothers Cobourg01 and Cobourg02 at significant amounts. We have termed these two as Unknown1 and Unknown2 or U1 and U2.

U1 matches C01 at 124 cM and C02 at 164 cM. U2 shares a total of 104 cM with C01 and 62 cM with C02. Additionally, U1 matches five others descended from John G. Owston with two of these not being part of this project. Those in the project are C07, C11, and C-S. He also matches CX3 who descends from Frances J. Owston.  Unknown2 matches three additional members of John G.’s progeny (C07 & C11) with one of those not in the project. C07 and C11 are sons of C02 and C-S is the son of C-R.

A 23andMe customer from this family, Unknown3, can also be compared to the Cobourg Owstons. Unknown3 (U3) shares the following amounts with members of our project (not counting the additional five children of C01, C02, and C03).


Between the three members of this other family that can be traced to the same ancestral couple, matches to five children of William and Frances occur.

Due to larger matching autosomal DNA, it appears that U1, U2, and U3 are probably descended from John G. Owston or his progeny. The connection to the other four other Owston siblings suggests that an Owston line relationship exists.  The three unknown relatives are related in the following manner:  U1 is the first cousin, once removed to both U2 and U3. U2 and U3 are first cousins.


The average match between U1 and C01 and C02 is 144 cM – suggests a half second cousin relationship – in the most likely scenario, they are half second cousins, once removed.  The average match size of 63.3 cM between U2 and C01 and C02 and between U3 and C01, C02, C03, C06, and C-H suggests a relationship of half third cousins, which is the most likely relationship.

Therefore, most probable scenario is that John G. Owston is the father of either the originating husband (H) or wife (W). Although, John G. Owston’s only genetic, legitimate son (Newton French Owston, 1854-1928) could have fathered the husband (H), he lived 20 miles west of Pittsburgh at the time of the husband’s conception. His father lived across the river in Allegheny City – 2.5 miles away. If N.F. Owston were the father, the amount of shared DNA would be larger.

N.F. Owston could have not fathered the wife (W), as he was living in Connecticut during the time of her conception.  His father lived 3.5 miles distant from the wife who was born in Allegheny City. Other scenarios have John G. Owston fathering the husband’s mother or the wife’s father between 1849 and 1851, as both were born in the Pittsburgh area; their spouses were born overseas. In either scenario, the amount of shared DNA would be smaller than it currently appears.

Since U3 is the only one from this family that can be compared to members of the Cobourg Owstons, there are several segments that are triangulated. To compare U3 to the descendants of John G. Owston, I created a second profile. While there are descendants of John G. Owston’s oldest child, Frances Jenett Owston (1852-1943), none have yet tested.  Therefore, I painted the matching segments between the three children of Newton French Owston who produced issue. These include the following:  C-H from MMO; C01, C02, and C03 from GHO; and C06 and C-R from OFO. I matched U3 to these individuals and looked for triangulation.

Since C-R tested with FTDNA, his matches to U3 won’t show, but they are implied. This is seen on Chromosome 17 of the second cousin profile.  Compare the magenta (MMO vs OFO) and the purple segments (U1 vs MMO).  C-H matches C-R in magenta and U3 matches C-H in purple. It is reasonable to believe that U3 also matches C-R; however, these two cannot be compared at the present. Additionally, C-R’s son, C-S, matches U1 on Ancestry with 103 cM.


What can be compared is a triangulated segment between U3 and two grandsons of GHO (C01 & C03) and a grandson of OFO (C06). There is a 17.4 cM shared match on Chromosome 10.



In addition, there is a potential triangulated segment between U3 and descendants of two brothers of John G. Owston. Returning to the DNA Painter original profile, note the mauve and light blue segments on Chromosome 2.  Fourth cousins C05 and C08 have a matching 8.1 cM segment (in mauve). C05 and U3 also have a match in that same area of 6.3 cM in light blue. Since C08 cannot be compared on 23andMe, as he tested with FTDNA, there is the probability that U3 and C08 also matches with 6.3 cM.


While I am confident that one of the grandparents of U1 and great-grandparents of U2 and U3 is the child of John G. Owston, it is impossible at this juncture to know which one it is. I have attempted to contact all three and U2 was the only one to respond. Like me, he assumed the connection to his family was through this couple. If it were the husband (H), this could be confirmed via Y-DNA testing.

Of these three individuals, only U1 is a direct male descendant. My next task is to contact U1 via the postal system and offer to provide Y-DNA testing to confirm or eliminate his grandfather as being my great-grandfather’s half-brother.

While there is an additional shared match between John G. Owston’s descendants and U1 and U2 on Ancestry, further examination of this match indicates that it comes to Newton French Owston’s descendants through his first wife’s (Merry Emma Merriman 1856-1895) family. U1 and U2 do not match the many other Merriman descendants who have tested on Ancestry and who are related to N.F. Owston’s children. U3 does not match any of the Owston’s Merriman family on 23andMe. It is believed that the matches to U1, U2, and U3’s family and to the Owstons are independent of each other.


DNA Painter is an excellent tool not only to track one’s own ancestry, but it can be used effectively in a descent study. If you are studying a pair of ancestors who are not too distant, I would highly recommend using DNA Painter in this regard.


Comments RSS
  1. Michael

    This was a fascinating read. It has me thinking about how I can apply your approaches to my own efforts to confirm whether my 5th great-grandfather had siblings (I have a couple candidates).

    Did you have trouble encouraging folks to test, and then upload to GEDMatch? I’ve found dozens of matches, but they’re all on Ancestry.

    • jowston

      Michael: Good luck on this, but a 5th great-grandparent may yield few matching results due to the distance of the matches. However, I have some matches that go back to a seventh great grandparent, but these are too sparse to match. If this is a male line match, you might have better luck with Y-DNA. As far as encouraging, some tested on their own; however, I paid for the kits in a number of instances. I have a few that will not upload to GEDMatch. There are five matching individuals on Ancestry that I cannot get to reply. In addition, there are three others that are connected to this mystery family, but I do not know how. A chromosome browser might help.

      • Michael

        Thanks for the input.This is helpful and I appreciate it. It’s my 5th great-grandfather, but a handful of his 3rd and 4th great-grandchildren have tested. It’s their tests that I’ve been using to identify matches mindful of the challenges you flag. Unfortunately, I’m trying to link him with a sister which limits the value of Y-DNA in this instance.

      • jowston

        Then you should have success with this venture then.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: