How Pairing Gworks with a Network Graph can Help Solve Your Research Objective

Gworks is a tool at DNAGedcom.com that quickly generates surname lists using information from your DNA matches’ trees. I have found this to be an efficient and valuable tool when seeking an unknown ancestor. When paired with a network graph it becomes very powerful.

Gworks and a Network graph work well together for the following types of objectives:

• When you know the surname of the research subject and you are trying to find his or her father. Quickly finding all matches with that surname in their tree can help you find possible ancestors quickly. Knowing a possible location for the family can help you narrow the search even further.
• When you have more than one paternal candidate and you know each of their wives’ names, you can look for the wife’s maiden name in your DNA matches’ trees using Gworks.
• When you have an unknown parentage case and you have identified a few candidates, identifying surnames related to the candidates and then searching for those surnames with Gworks can help determine whether there is stronger evidence for a genetic connection to one candidate over another.

 

Performing additional analysis with a network graph can help determine whether the matches you have discovered are relevant to your ancestor of interest. Here’s how it works.

Generating a Surname List Using Gworks

Gworks pairs well with a network graph because you will have already gathered the files you need using the DNAGedcom Client. For a refresher on how to gather these files, read Nicole’s article, Creating Gephi Network Graphs Part 1: Gather Matches and Prepare Spreadsheets. The data table created by Gephi is the tool that you will be using in conjunction with Gworks.

If you already have your network graph, you can next upload the files used to create the network graph and generate a surname list using Gworks. If you don’t already have a network graph, read Nicole’s series on Creating Gephi Network Graphs. Then come back here and follow these steps:

1. Locate the DNA Gedcom Client “m_” and “a_” files that you gathered for your Network Graph using the DNA Gedcom Client. You will want to use the original files, not the ones you adjusted to be used with Gephi.
2. Navigate to DNAGedcom.com and log in.
3. Select Gworks (Gedcom) > Upload DNAGedcom Gworks Files.
4. Use the select buttons to choose the correct files from your hard drive.
5. Once you have selected both files, click “Load the files.” In a few seconds you will get a message at the bottom of the screen that says “Finished Load.”
6. Click “Process the data.” In a few seconds you will get a message at the bottom of the screen that says “Completed Processing.

At this point, the data for your test kit is now loaded and ready for analysis. It’s important to note that Gworks will only store the data for one set of .csv files at a time. If you upload new files from a new test taker’s kit, the old files will be overwritten. You can always go back and upload the original files again.

Next, go back to the Gworks (Gedcom) tab at the top of the screen and choose “View Surname Occurrences.” This will generate a list of surnames that occur in your DNA matches’ trees. They are organized from most common to least common, and there are two columns: one for the number of files, and the other for the number of people with that surname among your matches.

Finding the Surname of Interest

First, search for the surname of interest in the Surname box at the top left of the screen. Type the surname and click the filter button next to the surname box. In this field, I usually choose the filter “Contains.” If there are known spelling variations for the surname, you can determine how much of the name to include in your search. For example, for a surname spelled “Hansen” or “Hanson,” consider searching for “Hans” and selecting the “Contains” filter.

Once the results for that surname are generated, click on the green arrow at the left of the name. This will open up each occurrence of that surname on your DNA matches’ trees. You can now use additional filters to help focus in on relevant DNA matches. If you know where the family might have lived, you can filter for that location in either the birth place or death place columns. You can choose a birth or death date and the filters “Greater Than” for “after” and “Less Than” for “Before.”

Checking to Make Sure the Matches are in the Same Genetic Network

Now that you have a list of DNA matches whose trees include the surname of interest, the next step is to use your network graph to determine whether the DNA matches are all in the same genetic network.
The most efficient way to check to see if the matches are in the same network is to download and use a csv file of the network data in the data laboratory in Gephi. Open the csv file and use CTRL + F to search for each match. Check the “modularity class” column. Modularity class is the term for the network or cluster number. If the matches are related to one another, they should all appear in the same modularity class or network.

If the matches are in the same network, it is possible that they all share a common ancestor. Belonging in a network previously determined to be connected to the family of interest provides additional evidence that the matches are connected to the ancestor of interest. Sometimes, the trees of these DNA matches are robust enough that they will show a common ancestor and you will simply need to verify the trees. If the trees aren’t very robust, you can build quick trees to try to find the common ancestor.

If the matches are scattered among different genetic networks, it can be assumed that they all share different common ancestors, and that the surname of interest is due to unconnected family members across the matches’ family trees.

Is the Person with the Surname of Interest a Direct Ancestor of the DNA Match?

While some DNA matches with the surname of interest on their trees might be in the same genetic network, another check is to determine whether the person with that surname is a direct ancestor of the DNA match. Gworks finds surnames for all persons on every DNA match’s tree, so the surname occurrence might be in a collateral line. If the person with the surname of interest is not a direct ancestor, that DNA match is not relevant for the study of that particular surname, even though he or she might be in the same genetic network as other matches on the list.

Is the DNA Match Related to the Research Subject Through a Different MRCA Couple?

Finally, even though a DNA match has the correct surname, is in the correct network, and does have a direct-line ancestor with the correct surname, you need to make one final check. Is the DNA match related to the research subject through a different MRCA couple? Analyzing all possible MRCA’s helps confirm or eliminate the surname of interest when using Gworks and a network graph.

Case Study: Confirming the parents of David R. Matheson of Pictou County, Nova Scotia

To illustrate these principles, I will share how I used Gworks to help confirm which of two Matheson men was the father of David R. Matheson of Pictou County, Nova Scotia. I had previously done extensive research in census and probate records to identify a number of paternal candidates and then eliminate them one by one. I was able to narrow the pool of candidates down to two men: Thomas Matheson of Rogers Hill, and David Matheson Esq. of Pictou.

Both Thomas and David had sons of the correct age in their households when the 1851 Census of Canada was enumerated.¹ No probate record was discovered for Thomas Matheson, and the children named in David Matheson’s will did not account for all males enumerated in his household in 1851.² Records revealed the maiden names of the wives of each of these men. Thomas Matheson married Jenette Clark, and David Matheson married Elizabeth McKay.³  Seeking Clark and McKay ancestors in DNA matches’ trees could reveal connections to either Jenette Clark or Elizabeth McKay, thus providing evidence favoring one candidate over the other.

Both Clark and McKay are common surnames, and Gworks revealed a fair number of occurrences for each. Because these families were second-generation residents of Nova Scotia, DNA matches that didn’t have the location of Nova Scotia as either a birth or death place were not examined in this analysis. Those who didn’t have the McKay or Clark surname on their direct line were also eliminated.

Only two matches with direct-line Clark ancestors in Nova Scotia were discovered. However, both were in the same genetic network, and an analysis of each of their trees revealed that their lines went back to the parents and grandparents of Jenette Clark. The matches did not share other MRCA’s.

There were more matches with the McKay surname in Nova Scotia in their trees, and several were in one of the genetic networks associated with the Matheson line. However, few of them had McKay as a direct-line ancestral surname, and those that did have direct-line McKay ancestors did not share the SAME McKay ancestors. Furthermore, most of them shared Thomas Matheson’s parents John Matheson and Jean Fraser as their common ancestors.

Having identified DNA matches with common ancestors on Jenette Clark’s line and not on Elizabeth McKay’s line, the evidence pointed to Thomas Matheson and Jenette Clark as the parents for David R. Matheson. Further documentary study of Thomas Matheson led to a deed created shortly before his death.⁴ The deed read like a will, with Thomas conveying his property to his son Kenneth, on condition that Kenneth fulfilled each clause written within the deed. One of the clauses provided “To DAVID MATHESON son of THOMAS MATHESON a feather bed and bedding when respectively demanded.”

DNA evidence pointed to Thomas Matheson and Jenette Clark as parents to research subject David R. Matheson. Documentary evidence solidified the conclusion that Thomas Matheson did indeed have a son named David, who was probably the research subject.

Summary

Gworks is a tool that can generate a list of surnames found on DNA matches’ trees. Seeking DNA matches whose trees include the surnames of interest, then using a network graph and additional analysis to determine whether the DNA matches are related can reveal evidence to help solve a variety of research questions. Backing up findings with documentary evidence can help solidify cases.

I hope that as you continue creating Gephi network graphs you will be able to think of ways to utilize the combined power of Gworks and your network graph to help you break down some of YOUR brick walls.

 

1. For Thomas Matheson, see 1851 Census of Canada, Nova Scotia, Pictou County, Pictou, line 15, household of *m?s Matheson; image Ancestry (https://www.ancestry.com/imageviewer/collections/1061/images/e096-e002385643 : accessed 9 November 2021) > image 23; Library and Archives Canada, Ottawa. For David Matheson Esq., see  “1851 Census of Canada East, Canada West, New Brunswick, and Nova Scotia,” Nova Scotia > Pictou County > Pictou > image 77, entry for David Matheson, Esq., database with images, Ancestry (https://www.ancestry.com/imageviewer/collections/1061/images/e096-e002385689 : accessed 16 September 2021); Nova Scotia Archives and Records Management (NSARM), Halifax.
2. “Canada, Nova Scotia Probate Records, 1760-1993,” images, FamilySearch (https://www.familysearch.org/ark:/61903/3:1:3QS7-89ZR-V9RV : accessed 9 November 2021) > Pictou > Index to estat, 1811-1940, vol 1-2, A-Z > image 210 of 297, negative search for Thomas Matheson. And “Canada, Nova Scotia Probate Records, 1760-1993,” images, FamilySearch (https://www.familysearch.org/ark:/61903/3:1:3QS7-L9ZR-V9W9-Y : accessed 11 November 2021) > Pictou > Estate files, 1886-1887, no 1962-1990 > images 156-171, specifically images 159-160; will of David Matheson Esq., written 27 July 1886, proved 6 October 1886.
3. Jenette’s maiden name and parents were discovered on her death record: “Canada, Nova Scotia Deaths, 1864-1877,” database with images, FamilySearch (https://www.familysearch.org/ark:/61903/3:1:3QS7-89JY-899 : accessed 3 November 2021) > Pictou > 1866-1868 > image 415; entry 4, Janet Matheson, 9 December 1868; Nova Scotia Provincial Archives, Halifax. Elizabeth (McKay) Matheson’s maiden name was discovered on two death certificates for her children: “Canada, Nova Scotia Deaths, 1890-1955,” database with images, FamilySearch (https://familysearch.org/ark:/61903/1:1:QK9S-J6WS : 8 March 2021) Elizabeth McKay in entry for Emily Matheson, 05 Mar 1922; citing Pictou Willau, Pictou, Nova Scotia, Canada, certificate 1468, Nova Scotia Archives, Halifax. And “Canada, Nova Scotia Deaths, 1890-1955,” database with images, FamilySearch (https://familysearch.org/ark:/61903/1:1:QK99-6F3D : 8 March 2021), Elizabeth Mc Kay in entry for Mary L Matheson, 12 Feb 1932; citing Pictou, Pictou, Nova Scotia, Canada, certificate 1042, Nova Scotia Archives, Halifax.
4. Pictou County, Nova Scotia, Registry of Deeds, “Deeds,” vol. 71 (1877), p. 59-61, Thomas Matheson to Kenneth I. Matheson, written 7 May 1875, recorded 21 March 1877; images, FamilySearch (https://www.familysearch.org/ark:/61903/3:1:3Q9M-C34L-3Z4D : accessed 5 May 2022) > FHL DGS #8621093 > images 35-36 of 398.