Evert-Jan Blom, who lives in the Netherlands with his wife and children had a background in Molecular Genetics before getting interested in genealogy. [i] After getting tired of manually trying to connect his matches into groups based on chromosome clustering, he wrote an algorithm (computer code) that did the clustering for him in record time.
This clustering has since become a very popular tool. In theory, every match on a specific cluster should ultimately share a common ancestor. While he is no longer allowed to run it against Ancestry.com® matches, his Genetic Affairs will run against your 23 and Me® or your FTDNA DNA matches. He has also written the same for MyHeritage® and Gedmatch®.
For example, on one of the clusters, I know, based on my own research, that three of the testers descend from the same family as my g-great-grandmother, Lurraine “Lizzy” Elizabeth Reese, the daughter of Samuel Reid and Sarah Catherine Reese, nee Sparks. Several of the Reese children married Akers and Hamiltons not to mention Akers marrying Hamiltons. So I can realistically assume that the others who share that cluster must also descend from those family combinations.
These are powerful tools and might be worth your time to investigate.
In addition, he recently added some reports that operate similar to Ancestry’s ThruLines which is based on what others have in their online trees. However, as the old axiom says, “Garbage in – Garbage out.” Hopefully, he will improve on this part of code which weighs too heavily on the same or similar name and not on timelines and locations. As I have said concerning ThruLines, if the result does not make sense, do not spend a lot of time chasing it.
As of today, any code that tries to predict pedigree and ancestry is basing the results on trees posted by others and the DNA which these modern testers have attributed to that line. I was hired awhile back to try and prove that a man who added the nickname Major (he never attained a rank higher than Sergeant) in front of his first name, William, was the son of a county Sheriff who also went by Major since the younger Major one got married in that county. I proved he was NOT the son because the sheriff’s son was William D., a lawyer who stayed in that county and always signed his name William D. Whereas the younger Major William moved to Macon, Georgia, and ran a munitions factory. They were obviously two different Williams. Plus, the sheriff had two other sons and a daughter.
However, numerous people who have researched the younger Major and his known brother have all attributed their father to the sheriff so ThruLines show the sheriff and makes it appear they share DNA with the sheriff. They do not. When clicking on the sheriff, all the DNA matches are via the two misattributed sons and zero sharing with the four known children of the sheriff.
Is this making sense? To rephrase, the sheriff had four known children, if the younger Major was a son, then his descendants and his known brother’s descendants would share DNA with the four proven children of the sheriff and there are zero matches. It gives a false narrative.
You can take the information as clues and that is about it. My oldest proven Thomas ancestor had four children and none are named Joseph. However, some have attribute Joseph Thomas, born circa 1818, to my known direct ancestor. However, the reason we do share DNA is that Joseph is actually a grandson of my ancestor and not a son.
That is not to say we cannot use it. I have had several clients move various men in and out of their 4th or 5th great grandfather position just to see the various results. We are looking for the highest shared cMs with cousins. Case in point, we are trying to find the father of a Joseph Bishop who was born about 1790, probably in the Spartanburg County area, and appears in the 1810 U. S. Census in Hall County, Georgia. He then moved to Campbell County, Georgia. In trying to determine the most logical Bishop family he descends from, we have moved a lot of potential fathers in and out of those positions. All of the Bishops are related to some extent. We have narrowed it down to two distinct Bishop families as the most likely.
This does not solve the problem but it does give us a good starting point to spend time researching closer.
Be careful what you believe. Be careful of any predicted pedigree.
[i] https://www.geneticaffairs.com/about.html