Your DNA Workshop
Welcome to the DNA Workshop! In every issue DNA adviser Karen Evans will be unpacking a reader case study, and giving you step-by-step tips for you to use in your DNA research at home. It’s time to unlock the story of your genes…
This month, Karen helps Family Tree reader Peter and follows the DNA trail of a family who have produced just fifteen descendants in 200 years, providing a great example of how to proceed in a case of few DNA matches.
PETER WRITES:
Last year you kindly helped me understand my ‘Y’ DNA test and featured it in Family Tree. I have now undertaken the Ancestry DNA test as they had an offer, and just received the results. I recognise some of the matches’ names though strangely not others in the 2nd-3rd cousins range. Once I have worked out how to do it, I will make contact with the more interesting ones (probably not bothering with the person with 30,000 people in their tree!)
However, my main interest is still in finding a match in my paternal Cope line. I am pretty sure that I have all the Copes, since Thomas Cope (born about 1810-15, not in Essex, died 1844). There are just fifteen of us: thirteen males and two females in 200 years!) I need to search, I think, amongst the poorer matches in the 4th+ cousins.
Do you agree and can you give me any advice as to the best way of approaching this?
Inputting Cope gave me a reasonable list of 5-8th generation plus matches. How can I pursue these further? If I contact the individuals, how do I get a reply?
KAREN REPLIES:
Peter has done extensive research on his Cope family, following all traditional documentation routes. He cannot get any further back than his 2x great-grandfather Thomas Cope. Could an autosomal test help further his research?
HOW CAN AUTOSOMAL TESTS HELP?
Autosomal tests are most useful at finding/confirming matches who share common ancestors up to five generations back. Researchers have found we should always match 2nd cousins or closer, but after that the chances we match a more distant cousin begin to fall. We have a 90% chance of matching a third cousin but only a 50% chance of matching
a fourth cousin. Other descendants of Thomas Cope (and any potential parents) are within five generations, so there is a chance of matches.
ASSESSING THE EVIDENCE
First things first. Did Peter’s DNA matches tally with his tree? Peter felt comfortable to attach a private tree to his DNA results which generated some Thrulines and common ancestor hints.
I was given access to view his results and could quickly see that his paper research was in line with his DNA matches. Peter has three 2nd cousin range matches which showed clear links to his father’s and mother’s maternal lines. He also has several high 3rd cousin range matches which confirmed his mother’s paternal line.
Peter’s grandfather was George Cope, born to James Cope and Emily Dennison. This was the only grandparent who we had not yet identified in DNA matches.
‘Val’, a 28cms match shows a common ancestor hint – she could be descended from Emily Dennison’s brother George. Using traditional research methods, we were able to establish that Val and Peter were 3rd cousins twice removed through the Dennison line. I clicked on Val and Peter’s shared matches.
SW shared 51cms with Peter but their tree was private. Peter contacted SW and they were able to establish they were 3rd cousins once removed, again through the Dennison line.
LACK OF DNA MATCHES – WHA T NEX T?
So far, so good. But no Cope DNA matches. I was also beginning to see Peter’s problem – lack of DNA lines to follow. Looking at census and baptism records it seems that the Cope family were not creators of large families! George Cope seemed to be the only child born to James and Emily, and when I talked to Peter it became clear that James was the only surviving son of Thomas Cope and his wife Mary Ann Legg/ Lake (there was also a sister Sarah). Without descendants there is a smaller pool of possible testers.
Peter has no idea who the parents of Thomas Cope were. Thomas married Mary Ann Legg/lake before civil registration, only appears in the 1841 Census to say he is not from Essex and dies in 1844 aged thirty (which ties into his age in the census).
Should Peter then look at the low Cope cousins in his DNA matches?
This is where danger lurks. You are almost always going to find matches whose trees contain your surnames but that is not necessarily at all where your shared ancestor is from. I have
a 4x great-grandmother who was a Jones before she married. Try putting that in the shared surname search! But even less common surnames will not ensure you are on the right trail. One of my great-grandmothers was named as Emma Hawkins on her marriage certificate. Eventually I found out that she was in fact born Emma Beavans but her mother later married a Mr Hawkins and Emma had taken his name. Evidence suggests Mr Hawkins was not Emma’s father and I don’t have any Hawkins ancestors. Yet, if I put that surname in a search, matches in the 5th-8th cousin range are found.
Peter has fewer than thirty matches whose trees contain the Cope surname and none of these matches are over 18cms. Of course, we are looking for distant shared ancestors so the centimorgans will be low but there is no way of knowing for sure if these matches are from the Cope line, another one entirely or even false matches. Peter could look at the shared matches of the Cope group to see if there are further clues on where they may belong in his tree. I know Peter is also going to see if there are any common ancestors among the Cope group – if they are all descended from the same Cope family it may be a potential lead. As we don’t know where Thomas Cope was born, we can’t be sure if the trees containing Copes from Essex, Staffordshire or even the US hold potential clues. I really advise caution though, it would be easy to jump to confirmation bias in this situation, it is extremely unlikely Peter can be sure he is descended from one of these lines (if any) or find evidence to support it.
NEXT STEPS
We must also consider whether George Cope’s father wasn’t James Cope. If George’s father was someone else then we need to examine any ‘mystery groups’ who didn’t fit into the known tree. The highest ‘unknown match’ shares 60cms with Peter and creates a small group, some with trees. Peter could look for common ancestors among this group. Was anyone belonging to this group in the right place and time to be a possible father? Perhaps some of Peter’s higher matches are connected through the Cope line but have no tree attached to their DNA results? They would not show up in a surname search so Peter could look to build ‘quick and dirty’ trees for these matches.
Peter could look for possible
THIS IS WHERE THE DANGER LURKS. YOU ARE ALMOST ALWAYS GOING TO FIND MATCHES WHOSE TREES CONTAIN YOUR SURNAMES BUT THAT IS NOT NECESSARILY A T ALL WHERE YOUR SHARED ANCES TOR IS FROM
descendants of James’ sister Sarah in his DNA matches. He knows she married and had several children. This may give him some leads to follow.
Does Peter’s DNA match any of the female lines who married into the Cope family? If, for instance he could find DNA matches to Mary Ann Legg’s line and open up his tree that way.
The more matches, the better. A really important potential cousin could have tested with another site so I would upload my DNA to other testing sites. If Peter is happy, after reading their terms and conditions, to upload his DNA to other sites he should certainly do so. Places like Myheritage, Familytreedna, Livingdna and Gedmatch allow uploads from other sites and have every growing databases. More matches could increase his chances of finding other possible descendants of the Cope line. 23andme do not accept uploads but it does have the second largest group of testers (after Ancestry) so Peter could also test there. Remember there are sales of DNA tests throughout the year, often connected to celebratory events like Christmas or Mother’s Day so it is well worth waiting.