Top 5 (and first 5) Tasks for atDNA Results

You’ve got your DNA test results back from Ancestry, FamilyTreeDNA, MyHeritage, 23&Me. You’ve looked at your ethnicity results and laughed. Now you know why you like guac so much, and why you love to the polka. And now you have a reason why you like cold weather much more than hot.

Or, the disappointment comes. Is that it? You’re faced with the reality that you’re exactly who you always thought you were. Nothing exotic in your past, but plain old, boring European.

Take heart, fearless test takers, there’s more to your results than “who are my people?” Those cousin matches are the keys to breaking through brick walls, uncovering new stories, meeting great people who love what you love, and shoring up what you always thought was true.

My husband is a data analyst at a large bank. I’m not actually sure that I can tell you exactly what he does everyday. It’s that kind of job. Truly, I try to pay attention when we chat over dinner. But, the acronyms are hard to keep track of! Whole conversations can take place without actually saying any English words. Amazing. But, in spite of the technical nature of his job, there are valuable lessons interspersed in there about data collection and data management.

The next tips will create a foundation of data analysis. This foundation will help with almost every other tool out there – Gedmatch utilities, DNA painter, visual phasing, GenomeMatePro, and DNAGedcom.

Grab a cup of your favorite beverage and get to work.

You’re going to start data analysis, phase 1: data collection.

Work your paper tree your 5th great grandparents. Then, work forwards.

Why 5th? That’s a far as your test results will go accurately. You should be very skeptical of matches further than that without good cause. Matches less than 10 cMs (centimorgans – a measurement of share DNA. The higher the # the closer the match in most cases, with notable exceptions) are most often false positives.

If you can’t get to your 5th greatgrandparents, awesome. You have your first DNA project. Congratulations.

But, you can get to you 2x great grandparents. That’s good too! Sit with those 2x’s and find all of their adult children, who they married and all of their children. Find those children and who they married and their children. Do it all until you reach about 1940. Yep, that’s a lot of work. Do it over and over again.

It will pay off. I can’t tell you how often, I’ve identified a match simply because I’ve worked my tree forward and recognize surnames and locations. It’s not always helpful to just know the surnames of your direct ancestors. Know all those associated families. It’s the FAN club for DNA. Sons-in-law and daughters-in-law and grandkids and great grandkids.

Identify your DNA matches, 3rd Cousin and Closer. 

You don’t need a name or an exact relationship. Just gather all the information you can.

Think about it this way. You have a woman in your tree whose surname, birthday and birthplace, you know. But, you’ve never found her parents. Think about what do you know about her parents? You know: Her father’s surname, her parents’ likely residence or at least that her mother was at a certain place at a certain time (her birthday!) You know her parents likely ages (over 14 and under 50 for a woman, 80 for a man). That’s a lot of information about people on your tree whose names you don’t know yet.

A match that’s 3rd cousin or closer isn’t a false match. They are your family. You’ll know, how much DNA you share with the match. If you’ve tested your family members, you’ll know how much DNA your match share with each of them (could be more and could be less.) Consider how this match may be related to other 3rd cousin or closer matches. You might label the match, maternal or paternal.

Label those 3rd Cousin or closer matches. 

Every testing company has the ability to take notes about a match. Gedmatch, which isn’t a testing company, doesn’t have that capacity, so try downloading matches to a spreadsheet.

You should include in the notes section items such as: the total centimorgans, the # of segments, the longest segment (if available.) Why? Those 3 factors (total cMs, # of segments and longest segment) are, in order, the 3 top factors in relationship identification.

Next, look at the matches haplogroup and note it, if possible. It can be a guide to include or exclude certain parts of your tree in a search for common relationship.

Check the test takers profile for information. Or the test administrators profile. Does the administrator have more than one kit under their name? Can you tell if the tester or admin has logged in recently or are they a one-time user? Have they included a gedmatch kit #, a personal email, residence?

Every detail is important.

Evaluate their online tree, if applicable. 

How complete is it? What kinds of details are available? Is there supporting documentation? Are there surnames your recognize? Familiar locations? These aren’t confirmations of relationship, just starting off points. Note all possible connections.

With enough information in a match’s very small tree, you might recreate their tree on your own and work the family backwards. See if anything starts to look familiar.

Create Groups. Group close matches together based on who matches whom. 

For example, you have a first cousin match that you can identify has being from your paternal side. It’s a reasonable (but not always accurate) assumption, that any other match that matches you and your paternal 1st cousin, is a paternal-side match.

Create a group for your paternal 1st cousin. Give it a name. Gather the common matches and label them with the group/1st cousin name. Who in this new group matches each other? Who doesn’t match each other?

Since your first cousin shares your common grandparents, subgroups may start to form that might be related to your paternal grandfather and another group to your paternal grandmother. One good match with a great tree, would be all it would take to label a subgroup further.

Here’s an example from my own tree.

This is a male test taker on that matches me.

  • His username gives an alias. It’s too generic to be an email prefix.
  • He has no tree, public or private associated with his test.
  • There’s no information on his profile page.
  • Ancestry notes that he became a member in April and last logged in that same month implying that he took a test for ethnicity only and only looked at his results once.
  • I share 62.8 cMs of DNA on 3 segments.
  • I also know that this match doesn’t match my mother.
  • He matches my sister at 95/5 and my brother at 59/5. He’s a solid PATERNAL match.
  • The match with my sister puts him at 3rd to 4th cousins rather than my estimate of 4th to 6th cousins.
  • He matches my paternal first cousin as well.
  • But, when analyzing other 1st cousin matches, he doesn’t fit into my paternal grandmother’s side matches. He’s likely a match on my paternal grandfather’s side. But, that’s a tentative observation that’s awaiting confirmation.
  • I’ve also noted that I’ve emailed him via Ancestry on a particular date.

That’s a lot of information about one guy that I’ll probably never hear from. Record all of that data. Take advantage of every bit of information. It will likely pay off later.

Why did I pick this match to share with you? I have a brick wall at my paternal line great grandfather. I can’t find any of James Morrissey’s family. See more information about James here. This match may be the key to finding James’ family, and I don’t know who he is. He’s unlikely to ever return my emails. But, I’ll keep my fingers crossed and hope, knowing that I did everything I could.

Continue Reading

DNA: Test All Your Siblings

I’m on a hunt for the parents of the first of our Morrisseys in the U.S. James (see his profile pages) was born in Abbeyleix, County Laois, Ireland in 1832. The only indication of his parents are on the back of his marriage record: Michael Morrissey and Catherine King. I’ve yet to uncover any siblings, or other blood relations in my extensive research on James, including online Irish records like Griffith’s or Catholic Parish Records.

James being my great grandfather, you’d think it would be a piece of cake. Nope. Nopedy, nope, nope. I think that part of the issue is age. My dad was 48 when I was born (I’m the youngest of 6 kids.) My dad was 3rd out of 6 kids.  My paternal grandfather (Ambrose) was 49 when my dad was born. James was 43 when Ambrose was born and Ambrose was 10th out of 11 kids. Each is old enough to be a grandparent when their kids were born. What’s 4 generations on paper, could be 7 generations for another family or in another’s DNA. Those 7 generations outside of the range of accuracy for an autosomal test.

Y-DNA testing (of my brother) has yet to offer any helpful insight.

Back to the title of my post: Test all of your siblings.

I’ve tested 2 of my 4 living siblings. So, 3 kids in the same family. My mom is still with us, so I’ve tested her. One of my resolutions for 2018 is to work all of those paternal matches. So, if someone doesn’t match mom, their labelled paternal. I add them to an excel spreadsheet an farm all the data I can. I’ve gotten through all of my 4th cousin matches on Ancestry for me, my brother and my sister.

Here are a few things I’ve noticed in crunching the data.

  1. Test taker “B.” My brother matches “B” at 106 cMs. My sister matches at only at 33 cMs. I don’t match this person at all! We have a confirmed paternal 1st cousin who matches “B.”
  2. Test Taker “J.” My brother matches at 92 cMs. My paternal 1st cousin matches “J.” Neither my sister, nor I match “J.”
  3. Our relationship with “P” was identified based on an online tree. We’re 3rd cousins, once removed. “P” matches me, my brother, my sister and my paternal 1st cousin. But at small amounts. 13 cMs, 24 cMs and 14 cMs respectively. While, these numbers fit into the 3C1R range in the Shared cM Project, they aren’t what I expected.

In each case, the lack of sibling tests would have left me without DNA confirmation of several lines on my tree. I wouldn’t have matches “B” or “J,” and “P” would have been such a small match that I probably would have overlooked it.

So, test everyone you can. And get those spreadsheets warmed up!

Continue Reading

NGS Conference – Day 3-4 #NGS2017GEN

I had a plan. In the middle of it, I changed it. Friday morning I decided that my brain was in danger of exploding, so I headed to a fun lecture to just relax and enjoy. And my head exploded anyway.

I had planned to hit up the well-known speakers and thought-generators. I did and they were great as expected. Let me tell you about the unexpected.

Lisa Louise Cooke, of Genealogy Gems, spoke about using a mash up of Google Earth Pro and genealogist’s favorite map collector, Dave Rumsey. Can I tell you how much I appreciate a woman who knows her tech? It was might have been the only lecture that I gave an all 5-star review. (I don’t hand them out often not because speakers aren’t doing a great job, but because all 5 stars aren’t helpful.) I also loved that whomever was running her powerpoint for her, was step by step alongside her fluidly thru the session. Great job being so well-rehearsed and prepped – much appreciated! This was the only session when I immediately went online afterward. Get the audio on this one. Or subscribe to her site and track down the tutorial. It was jaw-dropping.

My favorite track for the conference: African American.

For my own research, Tim Pinnick‘s session on “Reconstruction 101 for African Americans” gave great resources for my great, grandfather, James Morrissey, a Radical Republican representative to the Virginia Constitutional Convention of 1867-1868, but also gave a detailed history of the Reconstruction Era. Andre Kearns also did a great job recounting how he’s combined DNA and paper research to trace his family’s history in remarkable ways. Both of these sessions were also recorded.

I had hoped to make some new connections with other genealogists.  But, I’ve made some new friends in addition. The conference exceeds expectations!


Continue Reading