Last week, I wrote about how I utilized the Leeds Method in the article, The Leeds Method. What I didn’t say is that I was sizing up the Leeds Method for how I could use the technique to paint additional segments of my chromosomes.
The Leeds Method divides your matches into four groups, one attributable to each grandparent. That means those matches can be painted to your four sets of great-grandparents, assuming you can identify the maternal and paternal groups. Hint – Y and mitochondrial DNA matching or haplogroups may help if you have no better hints.
For genealogists who know who their grandparents are, testing close relatives and cousins is a must in order to be able to associate matches with your four grandparents’ lines.
Please note that the Leeds method generates hints for genealogists by grouping people according to common matches. We must further evaluate those matches by doing traditional genealogy and by looking for segments that triangulate. The Leeds method in conjunction with the actual match results at vendors, combined with DNAPainter helps us do just that.
Since I’ve been able to sort matches into maternal and paternal “sides” using the Leeds Method, which in essence parentally phases the matches, I can use DNAPainter to paint them. Here are my four articles I wrote about how to utilize DNAPainter.
DNAPainter – Chromosome Sudoku for Genetic Genealogy Addicts
DNAPainter – Touring the Chromosome Garden
DNAPainter – Mining Vendor Matches to Paint Your Chromosomes
Proving or Disproving a Half Sibling Relationship Using DNAPainter
Combining the Two Tools
DNAPainter has the potential to really utilize the Leeds Method results, other than Ancestry matches of course. Ancestry does not provide segment information. (Yes, I know, dead horse but I still can’t resist an occasional whack.)
You’re going to utilize your spreadsheet groupings to paint the DNA from each individual match at the vendors to DNAPainter.
On the spreadsheet, if these matches are from Family Tree DNA, MyHeritage, 23andMe or GedMatch, you’ll copy the matching segments from that vendor and paint those matching segments at DNAPainter. I explained how to do that in the articles about DNAPainter.
I do not use mass uploads to DNAPainter, because it’s impossible to assign those to different sides of your tree or ancestors. I paint individual matches, including information about the match and what I know about the history of the segment itself or associated ancestor.
I only paint segments that I can identify with certainty as maternal or paternal.
Pushing Back in Time
Based on which segments of identified ancestors the Leeds matches overlap with at DNAPainter, I can push that segment information further back in time. The blessing of this is that these Leeds matches may well fill in several blanks in my chromosome that are not yet painted by people with whom I share identified ancestors.
Even if your maternal and paternal grandparents are intermarried on each side, as long as they are not intermarried across your parental lines (meaning mother & father,) then the Leeds Method will work fine for painting. Even if you think you are attributing a segment to your paternal grandmother, for example, and the person actually matches through your paternal grandfather, you’ve still painted them on the correct chromosome – meaning your paternal chromosome. As you build up that chromosome with matches, you’ll see soon enough if you have 9 matches attributed to John Doe and one to Jane Smith, the Jane Smith match is likely incorrectly attributed, those two lines are somehow interrelated or it’s a false positive match.
Because I work with only fairly large Leeds matches – nothing below 30 cM, I sometimes receive a nice gift in terms of painting large previously unpainted segments – like the one on my mother’s side, below.
Look at this large green segment on chromosome 19 that I painted thanks to one of the Leeds matches, Harold. (Note that the two long blue and brown bars at the bottom of each chromosome are my ethnicity, not matches.) Another benefit is that if a Leeds match matches on already identified segments assigned to ancestors, I’ve just identified which ancestral lines I share with that match.
The green Ferverda side match to Roland through the Leeds Method aligns partially with a segment already known to descend from Jacob Lentz and Frederica Ruhle who were born in the 1780s. I’m related to Roland somehow through that line, and by just looking at his (redacted here) surname, I *think* I know how, even though he doesn’t have a tree online. How cool is that!
Important Notes for DNAPainter
Word of caution here. I would NOT paint anyone who falls into multiple match groups without being able to identify ancestors. Multiple match groups may indicate multiple ancestors, even if you aren’t aware of that.
Each segment has its own history, so it’s entirely possible that multiple match groups are accurate. It’s also possible that to some extent, especially with smaller segments, that matches by chance come into play. That’s why I only work with segments above 30 cM when using the Leeds method where I know I’m safe from chance matches. You can read about identical by descent (IBD) and identical by chance (IBC) matches here.
What a DNAPainter Leeds Match Means
It’s very important to label segments in DNAPainter with the fact that the source was through the Leeds Method.
These painted matches DO NOT MEAN that the match descends from the grandparent you are associating with the match.
It means that YOU inherited your common DNA with this match FROM that grandparent. It suggests that your match descends from one of the ancestors of this couple, or possibly from your great-grandparents, but you don’t necessarily share this great-grandparent couple with your match.
That’s different than the way I normally paint my chromosomes – meaning only when a specific common ancestor has been identified. For someone painted from matches NOT identified through the Leeds Method, if I know the person descends from a grandparent, I paint them to the great-grandparent couple. People painted through the Leeds Method don’t necessarily share that couple, but do share an ancestor of that couple.
When I paint using the Leeds method, I’m assigning the match to a set of great-grandparents because I can’t genealogically identify the common ancestor further upstream, so I’m letting genetics tell me which genealogical quadrant they fall into on my tree. With the Leeds Method, I can tell which grandparent I inherited that DNA through. In my normal DNAPainter methodology, I ONLY paint matches when I’ve identified the common ancestor – so Leeds Method matches would not previously have qualified.
I don’t mean to beat this to death and explain it several ways – but it’s really important to understand the difference and when looking back, understand why you painted what you did.
Labeling Leeds Match Painted Segments
Therefore, with Leeds Method match painting, I identify the match name as “John Doe FTDNA Leeds-Ferverda” which tells me the matches name (John Doe,) where they tested (FTDNA) and why I painted them (Ferverda column in my Leeds spreadsheet,) even though I don’t know for sure which ancestor we actually have in common. I paint them to the parents of my Ferverda grandfather. Not John Ferverda, my grandfather, but to his parents, Hiram Ferverda and Eva Miller. I know I received my matching DNA through one of them – I just don’t know which person of that couple yet.
However, looking at who else is assigned to that segment with an identified common ancestor will tell me where in my tree that segment originated – for me. We still don’t know where in your matches tree that segment originated.
“Match To” Issues
Lastly, if you happen to select a “match to” person to represent one of your grandparent matches that just happens to be descended from two grandparent lines, you’ve had your bad luck for the month. Remember, your “match to” person is the first person (closest match) that hasn’t yet been grouped, so you don’t really select them. If you realize you’re getting goofy results, stop and undo those results, then select the next candidate as your “match to” person.
At one vendor, when I selected the first person who hadn’t yet been grouped and used them for the red column which turned out to be Bolton, about half of them overlapped with Estes segments that I’ve already painted and confirmed from several sources. Obviously, there’s a problem someplace, and I’m guessing it just happens to be the luck of the draw with the “match to” person being descended from both lines. The lines both lived in the same county for generations. I need to redo that section with someone whose tree I know positively descends from the Bolton line and does NOT intersect with another of my lines. However, I was able to identify that this issue existed because I’ve already painted multiple ancestor-confirmed cousins who carry those same segments – and I know where they came from.
These tools are just that – tools and require some level of analytical skill and common sense. In other words, it’s a good idea to stay with larger matches and know when to say “uh-oh.” If it doesn’t feel right, don’t paint it.
Breaking Down Distant Brick Walls
I’m still thinking about how to use the Leeds Method, probably in combination with DNAPainter, to break down brick walls. My brick walls aren’t close in time. Most of them are several generations back and revolve around missing female surnames, missing records or ancestors appearing in a new location with no ability to connect them back to the location/family they left.
In essence, I would need to be able to isolate the people matching that most distant ancestor couple, then look for common surnames and ancestors within that match group. The DNAGedcom.com client which allows you to sort matches by surname might well be an integral piece of this puzzle/solution. I’ll have to spend some time to see how well this works.
Solving this puzzle would be entirely dependent on people uploading their trees.
If you have thoughts on how to use these tools to break down distant brick walls, or devise a methodology, please let me know.
And if you haven’t uploaded your tree, please do.
Would I Do The Leeds Method Again?
Absolutely, at least for the vendors who provide segment information.
I painted 8 new Leeds matches from Family Tree DNA on my Ferverda grandparent side which increased the number of painted segments at DNAPainter from 689 to 704, filled in a significant number of blank spaces on my chromosomes, and took my total % DNA painted from 60 to 61%. I added the rest of my Leeds hints from Family Tree DNA of 30 cM or over, and increased my painted segments to 734 and my percentage to 62% I know that 1 or 2% doesn’t sound like a very big increase, but it’s scientific progress.
It’s more difficult to increase the number of new segments after you’ve painted much of your genome because many segments overlap segments already painted. So, a 2% increase is well worth celebrating!
Having said that, I would love for the vendors to provide this type of clustering so I don’t have to. To date, Family Tree DNA is the only vendor who does any flavor of automatically bucketing results in this fashion – meaning paternal and maternal, which is half the battle. I would like to see them expand to the four grandparents from the maternal/paternal matching they provide today.
We’ve been asking Ancestry for enhanced tools for years. There’s no reason they couldn’t in essence do what Dana has done along with provide the DNAgedcom.com search functionality. And yes…I still desperately want a chromosome browser or at least segment information.
I will continue to utilize the Leeds Method, at least with vendors other than Ancestry because it allows me to incorporate the results with DNAPainter. It’s somehow ironic that I started out grouping the Ancestry results, but wound up realizing that the results from other vendors, specifically Family Tree DNA and MyHeritage are significantly more useful due to the segment data and combined tools.
Getting the Most Bang for Your Buck
If you tested at Ancestry or 23andMe, I would strongly encourage you to download your raw data file from both of these vendors and transfer to Family Tree DNA, MyHeritage and GedMatch to get the most out of your DNA tests. Here is the step-by-step guide for how to download your DNA from Ancestry.
The uploads to those three locations are free. All tools are free at MyHeritage until December 1, 2018 when they will begin charging for more advanced tools. The upload is free at Family Tree DNA and the advanced tools, including the chromosome browser, only require a $19 unlock.
Here is the step-by-step guide for uploading to MyHeritage and to Family Tree DNA. Fishing in every pond is critically important. You never know what you’re missing otherwise!
How many segments of your DNA can you paint using the Leeds Method in combination with DNA Painter?
I receive a small contribution when you click on some of the links to vendors in my articles. This does NOT increase the price you pay but helps me to keep the lights on and this informational blog free for everyone. Please click on the links in the articles or to the vendors below if you are purchasing products or DNA testing.
Thank you so much.
DNA Purchases and Free Transfers
- Family Tree DNA
- MyHeritage DNA only
- MyHeritage DNA plus Health
- MyHeritage FREE DNA file upload
- 23andMe Ancestry
- 23andMe Ancestry Plus Health
- Legacy Tree Genealogists for genealogy research
Good to know that we’re connected to House Lannister. I would have preferred House Stark but we knowwe can’t chose our ancestors.As always, your info is great.Howard
I’ve been using the “match-one-of-the-4-grandparents” method for a few months and discovered the descendants of my paternal (paternal line) 3xgreat aunt are also related to my father’s paternal grandmother. I even found where my dad’s paternal grandmother connected to the same ancestors. It means I have two different branches that match independently of my dad’s direct paternal line. This has helped me isolate about a dozen or more people on my dad’s side who DNA match the descendants of all my 3xgreat aunts and uncles on my dad’s direct paternal line. But I don’t know how they connect. I’ve seen their family trees and don’t see anything obvious other than they eventually trace back to old colonial Virginia. So now I’m beginning to suspect when my dad said his direct paternal line goes back to old Virginia it includes several families we already match including his and they are mixed up together because the founder effect.
I’m definitely going to implement the Leeds method to step up the rigor in my research. I only wish I could find the documents I’m looking for….
Which leads me to my next question. Given that I haven’t found the documents/records I need, just how bad is the document loss in Virginia before 1830. Especially the area around Richmond, VA and the entire region down to Surry and down to Halifax?
A lot was destroyed in the civil war. Google burned counties. Many counties do have partial records. The FamilySearch site shows records for each county available.
If Ancestry.com DNA results are uploaded to FTDNA or GEDMATCH, will that segment information be revealed? Thanks for a great blog, cuz!
Yes, depending. Ancestry V1 is entirely compatible with FTDNA. That’s May 2016 or earlier. V2 is partially compatible, meaning that your closer relationships will be found, but not more distant ones. At GedMatch, with the new Genesis system, they are in a bit of flux. I’d recommend uploading to both places, because that’s the only way you’ll ever get segment information unless you decide to retest at FTDNA.
Roberta do you have a excel file example w/ formatting available for download? I know it is simple enough to recreate by the screenshot in your post. However I wanted to see if you have an exact file so I can more effectively recruit family to use this method when sharing the post with them.
No, but let me think about this. The problem I have to ponder is that everyone’s spreadsheet will look different and many will have more columns than others.
Pingback: 2018 – The Year of the Segment | DNAeXplained – Genetic Genealogy
DNAPainter is brilliant and I have just subscribed to take advantage of the enhanced facilities but mainly to show how valuable the tool is to all of us.
I’ve painted cousin matches on my paternal side (and a couple of uncle/aunt matches). I had shown the DNA as coming from my Dad – which it did of course – but having read your article and watched Blaine’s video, I want to change that to the ancestral couple which would usually be my grandparents. Easy enough.
My grandparents were actually full siblings. This is beyond doubt so the DNA from them actually came from the relevant set of my great-grandparents.
So, am I safe to show the ancestral couple as my great-grandparents in DNAPainter, which would be the case with a couple of 2C matches in the same line?
Yes, because you actually know for sure where the DNA came from.
I did a Leeds Method spreadsheet to help an adoptee find his BF. I got 4 columns OK, but 3 of them must be from his mother. I had one that we don’t know where they come from and the trees (if any) are showing either private or only 3/5 people which is no help. Plus I got 4 other columns of people that didn’t contain many people and are unknown and again no trees. Did I maybe do something wrong? And why would I have 3 columns when I should only know the mother’s lines? I’m confussed.
It’s possible that his father is from a part if the world that has not participated in testing. Have you tried Generic Affairs clusters and auto trees?