DNA results are not always what they seem at first glance.
I was walking through the exhibit hall at the National Genealogical Society conference in Salt Lake City in 2010 when I came across Bennett Greenspan, President of Family Tree DNA, explaining the then-brand-new Family Finder autosomal DNA test. When he told me it could test DNA across genders — no more only being able to test a son of a son of a son or a daughter of a daughter of a daughter — I bought a test kit on the spot. I could practically retire on what I’ve spent on other family test kits since then… and the whole process has been one of the most exhilarating — and most frustrating — experiences I’ve had as a genealogist.
I’ve had enough successes that I understand the promise of autosomal testing.1 But I’ve also had what I personally consider to be more than my fair share of results showing a close relationship the other person and I can’t trace on paper. (I have such a bone to pick with the parents of my second great grandfather George Washington Cottrell (1821 Madison County, KY – 1891 Wichita County, TX), who didn’t even leave footprints as they passed through this world.)
I’ve learned that there are a lot of tools that can help us understand autosomal DNA results… and I’ve learned how very easy it can be to be misled by what looks — in some tools — like a very very close match indeed. Especially if you’re like me and only read the instructions after you’ve spun your wheels for a while…
FTDNA Family Finder chromosome browser chart
Case in point: take a look at this screen capture of some of my results in a tool FTDNA calls the chromosome browser. The first thing to keep in mind is that each of the colors represents an area on each of the listed chromosomes where my DNA results match another person’s DNA results. So what we really have on the chart is results for four people: me; my older brother (in orange); my uncle (in blue); and my first cousin (in green).
The second thing to keep in mind is that these are very close relatives on this chart. Each of these three people matches me very closely, and we share a TON of DNA. On average, siblings share as much as 50% of their DNA; an uncle and niece will share as much as 25% of their DNA; and first cousins will share as much as 12.5%.2 The vast majority of folks I match don’t light up the chromosome browser like a Christmas tree the way these very close relatives do.
The third — and most important — thing to keep in mind about this chart is that it really doesn’t show what it looks like it does. Yes, it shows that my brother, uncle and cousin are very closely related to me. What it does NOT show is that my brother, uncle and cousin are very closely related to each other.
Despite those huge overlapping areas where the orange and blue and green all line up so nicely and neatly on this chart, the fact is, my brother is not related to my uncle and my cousin. That’s because he and I are both children of our father, but we don’t have the same mother. And the uncle on this chart is my mother’s brother; the cousin is the daughter of my mother’s sister. While my mother’s family has been in America since before the Revolution,3 my brother’s ancestors on both sides didn’t hit U.S. soil until the 1880s or later. Our father and grandparents arrived at Ellis Island on 6 February 19254 and his mother’s grandparents were all born in Sweden.5 No cross-pollination between the two sides at all.
Even some of the overlaps for my maternal uncle and cousin turn out to be areas where I match both but they don’t match each other. That can happen because of the way a “match” is calculated. Every area sampled has two paired markers called alleles. I could be AG in those areas, my uncle could be AA in those areas and my cousin GG. Since only half of the pair has to match me to be called a match,6 my A would match my uncle and my G would match my cousin — and they wouldn’t match each other at all.
So take a lesson from somebody who still has to be told, routinely, to RTFM.7 The chromosome browser is a great tool. But, like everything else, it has limitations that are explained in the frequently asked questions on the Family Finder test. We may all be related as descendants of some ultimate Adam and Eve… but maybe not as closely as the pretty colored chart makes it seem.
SOURCES
- Judy G. Russell, “Autosomal DNA testing,” National Genealogical Society Magazine, October-December 2011, 38-43. ↩
- Ibid., chart, 41. ↩
- November 1680, order as to Nicholas Gentry, York County, Virginia, Deeds, Orders, Wills (1677 – 1684) 6: 268; York County Microfilm reel 3, Library of Virginia, Richmond. ↩
- Manifest, SS George Washington, Jan-Feb 1925, p. 59 (stamped), lines 4-6, Geissler family, 4; “New York Passenger Lists, 1820-1957,” digital images, Ancestry.com (http://www.ancestry.com : accessed 6 Jan 2012); citing National Archive microfilm publication T715, roll 3605. ↩
- See 1920 U.S. census, Cook County, Illinois, Chicago, population schedule, enumeration district (ED) 2285, p. 250A (stamped), dwelling 7, family 7, Edward F. Anderson household; digital image, Ancestry.com (http://www.ancestry.com : accessed 21 Jan 2012); citing National Archive microfilm publication T625, roll 318. ↩
- See generally ISOGG Wiki, “Identical By Descent segment” (http://www.isogg.org/wiki/Identical_By_Descent_segment : accessed 21 Jan 2012). ↩
- Wikipedia (http://www.wikipedia.com), “RTFM,” rev. 2 Jan 2012. ↩
That’s a very impressive chart, Judy. I wish I had a clue what I was looking at — or for — on it! I understand what you said, and I believe it, but I don’t know what I’m looking at. Are you the dark blue in the background on which the other colors are overlaid or??? Can you explain the chart a bit? Or should I just got get an autosomal DNA test done and RTFM myself?! 😉
You DO understand it, Barbara! You got it exactly right. The background represents all of my autosomal DNA that was sampled for the test. The colored bars show where the autosomal DNA of each of my relatives matches mine. Where there is no overlay color, it shows that my DNA doesn’t match any of the three others shown. The key thing to remember here is that all this shows is how each of them matches ME. It doesn’t (necessarily) show a thing about how they do and don’t match each other.
Thanks, Judy. Good to know I have a small clue at least. But since they have each had their autosomal DNA test done, can you (or they) not create similar overlays to show where their DNA overlaps with each other? Or am I misunderstanding (a likely possibility).
You can create this kind of an overlay with each person who shows up on your match list (and people are only on your match list if you share enough DNA in common to reach a threshold designed to reduce the number of false positives). So I can do overlays with everyone who matches me, and — for example — my uncle and my cousin (who do show up on each other’s match lists since they are uncle and niece) can do overlays with each other. And the precise data represented by this overlay chart can also be downloaded for use in a spreadsheet.
That’s why you only need to worry about being misled when you compare people on your match list without considering if they show up on each other’s match lists. Neither my brother nor my uncle shows up as a match on the other’s list — it’s only on mine that the potential for trouble exists (and only if… sigh… I don’t read the instructions!).
I’ve submitted DNA from two male relatives to Family Tree DNA so far, with one getting me to the exact county in SC that the family was from. I have never seen this chart before and have no idea how to access it, but reading your post, I’m highly unlikely to ever even try. I think it’s a pretty picture and you understand why it has a “Christmas tree” effect. How many people are looking at these and wondering what it all means? What the point was of Family Tree DNA creating this option in the first place?
Rondina, remember that this is ONLY for the Family Finder test, and not for the YDNA (male only) test. So you won’t see this option at all if you haven’t had their DNA tested as part of the more expensive Family Finder option. The big difference between Family Finder on one hand and either YDNA (male only) or mtDNA (female only — with certain exceptions) on the other hand is that, with Family Finder, you can test across genders. And that means you generally DON’T share a surname with your matches. So FTDNA is trying to provide tools to help you and your matches narrow down where in your paper-trail family trees you should look for a possible common ancestor. Used properly (she says, with a red face, since it’s my own experience in NOT reading the instructions that lead to the post), the chromosome browser can be helpful for that. As long as you remember that you START by seeing where folks may line up on the pretty colored table (and don’t STOP there), it’s a useful tool. If you DO stop there, it’s a little bit like using a hammer when you really need a screwdriver.
Ah, the hammer for screwdriver approach. I know it well! As another who is afflicted with SRS (selective reading syndrome), I usually read things AFTER I’ve screwed up — using a hammer, of course — and now need to go back and figure out what I did wrong. I like the autosomal DNA notion, and I haven’t done it yet. I’ve done a full mtDNA sequence, have a third cousin who did an mtDNA level 1 test to help try to figure out where a certain family came from maternally (Finland? Really?) and have two relatives who have done Y-DNA tests which have helped by proving a connection in one instance and proving a non-connection in the other. So I’m fond of DNA even though I never understand all the columns of numbers and letters. Someday maybe it will be able to be discussed in layperson’s terms. Until then, I fear I’ll be admiring the Christmas tree.
“Someday maybe it will be able to be discussed in layperson’s terms.” Not to toot my own horn on this, but I think SOME of the basic questions you may have would be answered in the article I wrote for the Oct-Dec 2011 National Genealogical Society Magazine…
Thank you for an interesting perspecitve on Family Finder. I just received my Family Finder profile last week and have been playing with the “chromosome browser”. Your blog has helped me better to understand the false positives. After looking at the results of your brother and his non relatives, it makes me wonder just how useful it is to even look at the chromosome browser.
Tom, it really IS a help but it’s just the first step. The next thing to do is look at the results in more depth in a spreadsheet. I’ll be writing about that this coming Sunday (now that I’ve decided to make DNA the subject of most, if not all, Sunday posts from now on…)!
Great post Judy. And yes, this type of illustration does only show how people match you. I am most familiar with 23andMe which has a separate tool for showing how people match to each other. Does FTDNA offer the same?
Thanks, Taneya! Yes, FTDNA has a number of different tools that generally parallel those offered by 23andMe. I’ve personally tested with both companies and personally find the results at FTDNA far far easier to work with and use for genealogy purposes.
Hello, If I wanted to find out if my niece’s children are related to me, would this type of DNA test help?
Yes, this test (autosomal DNA testing) would be able to tell you if your niece’s children (your great nieces and nephews) are related to you, but you would have to be able to test them as well as take the test yourself. If they’re children, permission would have to come from their parents; you don’t have legal authority to act unless you’re their guardian.