Not a magic bullet!

I could virtually hear the shrieks of joy yesterday as social media exploded in response to Ancestry’s announcement of product updates, including new tools for working with DNA match data.

One of those tools, ThruLines™, uses data from users’ public or private, searchable trees linked (to their DNA results), and data from other public and private but searchable trees to show how users could connect to their DNA matches through common ancestors. Like two other new tools, MyTreeTags™and New & Improved DNA Matches, it’s still in beta.

ThruLines™ is a potentially powerful new tool, but users will get better results by understanding that it’s not a magic bullet.

Ancestry cautions users that suggested ancestors are only as accurate as the trees data is pulled from.

“Since ThruLines are based on the family trees of you and other members of Ancestry, they’re as accurate as the trees they’re based on.”Ancestry.com

The new suggested ancestors are just new information until users apply it to a specific research question. So, like all information, it could be true or false. Information doesn’t become evidence until it’s applied to specific research questions.

Since genealogy is all about research, it means that evidence needs to be documented, cited and analyzed before making conclusions, and conflicting evidence needs to be resolved.

Marc McDermott, with contributions from Elizabeth Shown Mills, CG, CGL, FASG and Thomas W. Jones, Ph.D, CG, CGL, FASG, has graciously shared this great graphic, via Creative Commons Attribution-NoDerivitaves 4.0 International (CC ByND 4.0): https://creativecommons.org/licenses/by-nd/4.0/. It shows what our research process may look like, even when working with new information from ThruLines™, or any other new info.

Genealogical-Research-Process
Genealogical Research Process – Marc McDermott, with contributions from Elizabeth Shown Mills, CG, CGL, FASG and Thomas W. Jones, Ph.D, CG, CGL, FASG.

This means users need to treat new information from ThruLines™ just like any other information!

An Example – 

ThruLines Pettit
ThruLines™ suggested common ancestor, Sarah Morris, for Melanie J. Rice, at Ancestry.com.

In the image above, I see two potential new direct ancestors, who are not in my tree. I was actually aware this could be my line of descent before seeing ThruLines suggestion, but I have not proven my more recent direct ancestor’s (William W. Pettit) connection to Allen Pettit, nor Allen Pettit’s connection to Sarah Morris, so they are not in my tree. I also had no idea I might connect to this many potential descendants!

Pettit Descent
ThruLines™ suggested line of descent for Melanie J. Rice from Allen Pettit, born about 1805, in Morgan County, Ohio.

Should I add them to my tree now that I see more than 40 DNA matches have them in their trees? NO!

I’m sure excited to learn that I seem to share DNA with several distinct branches of others descended from potential common ancestors! And, this discovery highlights an avenue for further research for me.

My plan going forward with this information will include research questions and evaluation of evidence. I may start with something like this:

Do each of the 15 DNA matches shown actually descend from the Allen Pettit, who was born about 1805, in Morgan County, Ohio? If so, how much DNA do I share with them? Do I match any of these folks through a different ancestor?

I may explore similar questions for the other DNA matches whose trees show they descend from other children of Allen Pettit. Next I would look for other evidence that may or may NOT connect us to Allen Pettit, and then evaluate all of the evidence before making a conclusion.

While I’m thrilled that Ancestry has provided users some great new tools, I wish they would prominently display a strong disclaimer, warning those new to genealogy that these are suggestions, not facts, and as such need to be verified.

In summary – 

Ancestry has given us some powerful new tools, including ThruLines™. We need to resist the temptation to take shortcuts around solid research methodology, and treat this information like all other new genealogical information.

To leave feedback for Ancestry on ThruLines™, go to Explore ThruLines™ on your main DNA page, select any of the potential ancestors, and click the “provide feedback” link in the dialog box at the bottom of the page.

Resources – 

Ancestry tutorial on new features

More from Ancestry on ThruLines™

TheDNAGeek’s blog post on ThruLines™

Evaluating Evidence

Evaluating Evidence in Genealogy Research

Genealogical Evidence, Sources and Information

Genealogy Explained workflow chart and GPS Standard

Advertisements

When Worlds Collide

I wasn’t expecting any exciting discoveries when I clicked on one of my new Ancestry matches this morning.

We shared just 66 cMs, and she had no tree. What could I really hope to learn from looking at this?

Surprise Cousin

I scrolled down this new matches’ shared match list. Wait. No. That can’t be right!

After seeing her first two matches, my brain lumped her into the paternal match bin. They were known close paternal cousins. But that third match – it was my maternal 2C, R.J.1, whom I’d just helped find his birth father. I know how he connects to me. He’s a maternal second cousin, descended from my great-grandparents, Francis Alvin VanLue and Elizabeth Jane Pettit’s son, Harry VanLue and his second wife, Dorothy Ritchey.

My Tree2

I doubled back. Neither my mom nor my sister, who have both tested at Ancestry, match this person.

But – R.J.’ s a proven maternal second cousin. Multiple shared matches and target DNA testing prove this. Could R.J. be a double match, with shared DNA from my dad’s side as well? Hmmm.

I drilled down deeper into new match, S.J.’s, shared match list with me. There were several other potential maternal matches, but they all linked to the same couple R.J. is descended from, and share just his paternal grandmother’s line. There were more of my paternal matches, first on my Gorsline/Clem branch, then, as the shared cMs became fewer, the shared matches trailed off into just Dailey/Majors, a line on my dad’s father’s side.

DaileyMajors

I was gobsmacked. If anyone told me R.J. and I connected on one of my paternal lines, too, I would never have guessed this particular branch. But indeed, it appears we very likely do share DNA from one of my ancestors on this line, in addition to sharing my mom’s grandparents’ ancestry.

I’m able to identify these connections quickly, because I use several tools to manage my Ancestry matches’ data. I use the MedBetterDNA Chrome extension to display my match notes on my main DNA match pages. I’ve coded my matches with colorful emojis that help me see at a glance how that match connects to me, if I have that information. And, I review all my 4C and closer matches.

MedBetterNotes

I also keep a spreadsheet matrix of all my 4C and closer matches at Ancestry. The sheet shows who matches whom in vivid colors that make it easy to distinguish maternal and paternal matches at a glance.

Matrix Snip

I used Heirloom Software’s DNA Match Manager tool to export a CSV file, with my Ancestry matches’ user names. I built my matrix spreadsheet using this data.

And, I use DNA Painter to map chromosomes when I know what segments I match particular cousins on.

Or, where I THINK I match them.

This new match changes everything.

DNA Painter

It means that segments I’ve mapped for my second cousin, R.J. are not necessarily what I thought they were. Since this match shows that he and I match on BOTH my maternal and paternal lines, I can no longer “assume” that segments I share with him are from a maternal ancestor!

This underscores Blaine Bettinger’s admonition to always compare complete trees before making genealogical conclusions. He’s shared his thoughts on tree completeness frequently in a Facebook group I belong to. I heartily agree with him. I practice this. I practice it religiously.

At least I thought I did.
My match with R.J. seemed like a “no brainer.” He shared DNA with a host of other cousins descended from my overachiever great-grandparents, Francis A. VanLue and Elizabeth Jane Pettit. Of course we shared maternal DNA.

We focused on finding his father, and in the process did identify his paternal grandmother. I pushed her line back a generation or so, to help when we sorted R.J.’s matches by whom they descended from. But that’s where I stopped. We found our match. We knew how we connected.

Or did we?
Turns out, we didn’t know all the ways we connected, and I now have more research to do, to decipher how his maternal grandmother’s line intersects my dad’s ancestry. And, I need to rethink my DNA Painter activities. For now, I’ll just keep this in mind, and watch for new match data to enlighten us on where we share the DNA from my dad’s side. (I don’t have access to my mom’s test results.)

And, I will be even more vigilant in comparing complete trees with my matches, and in documenting how I evaluated our match results, including how I compared our trees, searching for common ancestors.

Postscript:
I’ve decided to revive my Grandma’s Genes blog as part of Amy Johnson Crow’s 52 Ancestors in 52 Weeks Challenge
Though my research practices have improved dramatically over the years, I’ve done little to get my work “shareable” with family and the genealogy community at large. I hope working through Amy’s challenge will help me find a good format for sharing my work while challenging me to take my work to the next level.

I intentionally plan to give myself liberal leeway to skip here and there, as I already know life will get in the way.

Though this entry doesn’t exactly focus on a specific ancestor, it’s apropos for this week’s prompt: Surprise

1Names and initials have been changed to protect the identities of living people.