Ancestry’s New “Amount of Shared DNA” – What Does It Really Mean?

Yesterday, Ancestry quietly introduced a new feature of their AncestryDNA autosomal product called “Amount of Shared DNA.”

This can be seen when you view your match, beside the confidence bar, as shown below.  Fly over the little “i.”

shared dna

It’s nice to know how much DNA we share and across how many DNA segments – but what does this really mean, how is it calculated, and how do these calculations stack up against the same information from other vendors?

Why would it be any different, you ask?

Because Ancestry runs their academic phasing program, Timber, and removes segments identified as matching to many people, constituting pileup areas.  Remember when Timber was introduced and people lost more than half of their matches?  I went from 13,500 to 3,350.  Today, 50 weeks later, I have about 6,700.

Real phasing is when you utilize your parents DNA to divide your own DNA into half.  Half your matches match you and your mother, and half your matches match you and your father.  If not, then they are not IBD matches.

Timber attempts to remove segments that are too matchy – areas where Ancestry feels you have too many matches so they might be “population” based match segments instead of real genealogical segments.

This new “Amount of Shared DNA” feature gives us the opportunity to test their matching against other vendors.

Thankfully, my cousin Harold has tested at all the vendors and uploaded to GedMatch, as have I.

Therefore, we can compare our results on all platforms.

shared dna 2

Why is the Ancestry total cM so much smaller than the other vendors, at any threshold?  Timber.  Ancestry is removing many segments that other vendors are counting and using, even at higher thresholds like 10 cM.  In fact, at GedMatch, their maximum threshold is 10cM and even at that level, the total match cM was 135, 21 more than Ancestry, and the SNPs were all well over 1000.

shared dna 3

The Acid Test

I’ve believed since the introduction of Timber that it removed too many segments – segments that are valid and useful – thereby removing valid matches.

However, the acid test is a parent/child match.  Each child should match their parents on exactly 23 segments (or 22 if Ancestry is not counting the X chromosome), one complete match for each chromosome.  Once in a while you’ll have a read error that may divide a chromosome into two match segments, so an occasional 24 or 25 wouldn’t be surprising.

What are we seeing?  A quick read of forums and looking at the results I have access to shows me that parent match segments are ranging from about 85 to about 110, which, in case you are counting, is from 64 to 87 more than the 22 (or 23 counting the X) chromosomes that we have.

What this tells us is twofold:

  1. Timber is removing 64 to 87 VALID segments in parent/child matching, believing that pileups are invalid. Rule #1 of DNA – you must match your parents. If you double this number, because you have two parents, each person has in the ballpark of from 130 to about 200 areas where their DNA is “too matchy” and segments/matches are removed. This illustrates the magnitude of the Timber problem.
  2. You cannot draw or correlate any relationship inferences from either the total amount of shared DNA nor the number of segments by utilizing the typical tools utilized by genetic genealogists because Ancestry’s totals will be lower and their segments will be broken into more pieces due to the removal of segments identified by Timber as invalid matches.  Blaine Bettinger is beginning to collect information at this link on Ancestry’s shared cM data for known relatives.  This information will be made public for all to utilize, as has his earlier shared cM work.  Please contribute if you can.

Hopefully Ancestry will take this opportunity to address the Timber issue, and hopefully they will eventually provide a chromosome browser type tool.  Now all we need is the chromosome number and start/end addresses for those chopped up segments.  These tidbits and pieces of solutions are not appeasing the genetic genealogy community and this new “amount of shared DNA” feature will not “do” in place of a chromosome browser.  I know this sounds like a broken record…and it is.  While Ancestry seems to be inching in the chromosome browser direction by providing additional information….I wouldn’t hold my breath.  I don’t think it will ever happen – but I would really, REALLY like for Ancestry to prove me wrong!

Fortunately, Ancestry’s tree matches and Circles are useful and thankfully, we can download our autosomal DNA results to both Family Tree DNA and to GedMatch and utilize their chromosome browsers and other tools.  Unfortunately, not everyone is willing to download, so we do really need that chromosome browser.

43 thoughts on “Ancestry’s New “Amount of Shared DNA” – What Does It Really Mean?

  1. I concur Roberta. Especially with the impending changes at 23andMe, Ancestry has the opportunity to move to the forefront in DNA genealogy testing, if they would simply add a chromosome browser tool.

    • Why won’t Ancestry DNA add a chromosome browser? And six months ag ftdna refused ny ancestry upload saying it was misaligned. Yet GEDMATCH and My Heritage had no problem with it. Answers to these questions would be most appreciated.

      • I suspect Ancestry won’t add a browser because they don’t have to in order to keep selling kits. Business decision, plain and simple. Try your upload again.

  2. My daughter and I share 3,405 cM across 76 segments according to ancestry. On gedmatch we share 3,582 cM with the largest being 281.5 and 196.1 on the x chromosome.

  3. I have had a lot of experience with Ancestry.com, doing my family genealogy for 20 years. I am pretty sure what is going on, this is the Mormon church and there doctrine says they are the original Native Americans which is an out right lie. Their DNA doesn’t support this, so this is why all the confusion. They are trying to do away with the Real Native DNA and insert their DNA to support their Doctrine. For example two of my cousins are as full blood Native as can be 100% according to Ancestry. When their son and daughter tested according to Ancestry the son and daughter were only 14% ? And ancestry made no connection with their parents or the brother and the sister ? I think this is what is happening. I am also a Biblical scholar, Hebrew and Christian and a Political Scholar. I have been studying these topics for over 30 years. John Provencio Ortega Harding

  4. I had the DNAgedcom tool so I already knew the Ancestry cM totals – spot on, nice job guys – but I was quite surprised to see Ancestry had MORE segments in some cases than either GEDmatch or 23andMe for those same cousins. If they can show some tiny segments for close cousin matches are IBD that would be valuable.

    A week ago I’d have said there’s almost no chance Ancestry would adopt a chromosome browser. They’d be worried about a few loud customers complaining about privacy (needlessly IMO) and the customer support. But now – I’d say the odds have increased just a little. If you give customers the keys, the brakes and now the power windows – sooner or later you just might give them the steering wheel.

      • I can definitely vouch for this. My wife, our daughter, and I have results from both 23andMe and Ancestry.

        At 23andMe, my daughter shares 3718 cM with me across 26 segments, and 3718 cM with her mother across 25 segments. It should be only 23 segments in both cases, but one or two false “breaks” isn’t surprising.

        At Ancestry, our respective sharing is “3,468 centimorgans shared across 54 DNA segments” and “3,455 centimorgans shared across 60 DNA segments”.

        That’s over double the number of segments in each case, but at least 250 fewer centimorgans. If anyone doesn’t think 250 centimorgans isn’t a lot, consider that it’s greater than the average amount of DNA shared between 2nd cousins (212.50 cM).

  5. Roberta,
    Why do think AncestryDNA does not provide a chromosome browser? I would think doing so would make them much more competitive and would increase their market share.

    • My personal opinion is because it would increase their support staff load to explain and educate people on what the results mean. And, they obviously don’t have to in order to sell over a million kits. That is only my personal opinion and is not something Ancestry said.

    • I agree with Roberta on this, but I would add that Ancestry’s primary source of growth (and main DNA test revenue source at this point) is people new to genealogy and probably very non-technical. These people don’t want to learn about centimorgans and segments and triangulation and IBS/IBD. So Ancestry has created a nice self-serving rationalization by arguing triangulation and segment analysis isn’t the right way to do things and they have provided “better” tools (circles and NADs) that don’t require a chromosome browser. They also are very anxious to provide a highly “successful” experience for their customers (whether it is accurate may not make much difference to the unschooled – just look at their advertising bragging about how easy it is to do genealogy.) So they want to do whatever they can to provide MORE matches, MORE shaky leaf hints, MORE trees, etc. whether or not the underlying process is producing accurate results. Segment analysis might expose all the false matches and false matches of matches (reducing the number of people in circles). So it’s bad for business. (Just my humble guess)

      That said, I get more from having tested at Ancestry than I do from my tests at FTDNA and 23andMe. Having thousands of matches (even weak ones) and far more trees (even sometimes inaccurate ones) provides more clues to sift through. I like all the babies in my bathwater and I’m willing to put up with more bathwater to have that.

    • Dr. Cathy Ball at AncestryDNA does NOT believe that segment triangulation is valid science. She believes that the probability of two descendants inheriting a shared segment of DNA is very low.

  6. I have TWO big issues with this new AncestryDNA feature. A known papertrail second cousin and I share 39cM across five segments and THREE known papertrail fourth cousins come in at 64cMs across five segments. Without a browser we can’t see if those five chromosomes are the same! I’m working on these cousins, slowly trying to talk them into transferring to GEDMatch for some serious study!

  7. Thank you for the heads up!

    I had just gotten the DNAGedcom Client tool earlier this week, so I also had the numbers already, but it’s nice to see them directly from Ancestry. It’s a step in the right direction and I expect/hope more will follow.

    As for the low totals and high segment counts, yeah, that’s a problem. Total numbers for lower level relationships tend to be too variable to precisely indicate relationship level anyway.
    However, since the totals are consistently low, they can still indicate a baseline and help differentiate between very close relationships as the cM differences between these high-level relationship levels is large.

    For example, with a total of 1800+cM I was finally able to confirm a half-sibling relationship rather than 1st cousin for the mystery match at the top of my list, and that is beyond huge. Similarly, a handful of other totals were clear enough to confirm a few additional very close relationships (some of which had already been confirmed on GEDMatch). The numbers are probably good enough down to 2nd cousin level at least and maybe a step or two beyond. For those who are unraveling recent mysteries and have very close matches, this is invaluable.

    That said, I saw some folks with totals way, way below my GEDMatch values. And yes, I’m still yearning for the chromosome browser.

    Love the blog,
    Chris

  8. Roberta, thanks for your consistently great information. Yes, my son has 95 segments with me and 95 with his dad. We have also seen some valid matches from GEDmatch that came from Ancestry but don’t who up on Ancestry at all in the 10 – 20+ cMs range. One person who is adopted matches 3 sibling tests that I manage on GEDmatch but they don’t show any match at all on Ancestry with 20+ cMs on each test even though they all tested at Ancestry. Very hard to help that person

  9. In addition to the magnitude of segments cut, I’m very skeptical now of their 5th-8th cousins. One in particular I’d favorited as a possible distant cousin showed a match of only 5.2 cM over 2 segments. I imagine this means each segment is probably no bigger than 3 cM which is ridiculously small! It makes me feel like they’ve decided to supplement the loss of matches due to Timber with truly irrelevant matches.

  10. The Ancestry download process is confusing. I have 3 cousins who can’t manage it (all professionals now retired who must have had someone else operate their computer. I always recommend FTDNA but ancestry’s cheap price and advertising trump that. I’ve started sending ftdna kits now.

  11. This was fascinating! So, I had to check my dad and me on Ancestry… we share whopping 113 segments on Ancestry!!! Sheesh! (It’s only 23 segments on FTDNA, naturally.) Probably giving me false positives and false negatives on some of my paternal cousins.

  12. Frankly, I’m surprised that AncestryDNA has given us this much information about our matching segments. It makes the quirks of Timber so obvious. They really have nothing to lose now by going all the way with complete matching segment details. What are they waiting for?

    • I have a New Ancestor Discovery in AncestryDNA where I have been “roped into” a DNA circle of people descended from a Richard Heavin (1770-1860). He and his father were born in VA, which is where I have MANY ancestors, some born early in the 17th century. While poking around one of these DNA cousins (14.1cM on one segment) I found a possible connection to my Eassery family line, but they have no other info on that line other than dates. It is quite possible these Heavin descendants and I have a common ancestor from VA or MD, but my guess the link is in a female line where female lines get lost very quickly.

  13. At Gedmatch
    My mother and I share 3,569.4 cM
    (segments at least 7 cM in length)

    At AncestryDNA,
    my mother and I share 3,305 cM across 101 segments!

    That is a 264 cM difference.

    My father has been dead for over 20 years, and so there is no way of checking my DNA against his.

    but my paternal Aunt Carrie tested
    she is my father’s maternal halfsister, and so she’s my half aunt

    At Gedmatch,
    my paternal Aunt Carrie and I share 912.3 cM
    (segments at least 7 cM in length)

    At Ancestry DNA,
    my paternal Aunt Carrie and I share 766 cM

    That is a 146 cM difference.

    I agree with Roberta!
    AncestryDNA is removing a lot of valid segments!
    I don’t trust their Timber algorithm.
    I never did after I started losing matches that matched both my mother and me at other sites.
    I lost paternal Acadian DNA segment matches too, and some of them match me on other sites.

    I also checked my largest Acadian segment match at Ancestry. It’s a Cajun woman that matches both me and my Aunt Carrie with a 27 cM segment match on Chromosome 5 but appears as a match to only me at AncestryDNA. AncestryDNA shows only 11.6 cM across 1 segment. That’s 16 cM less than at Gedmatch.
    I am thinking that AncestryDNA wiped out much of my paternal European Chromosome 5 segment. I had three 11 cM Chromosome 5 matches that are 3 generations directly related at Gedmatch that used to appear at AncestryDNA. It was in the same area where I am matching people with Acadian ancestry.
    At Gedmatch, my largest chromosome segment with my Aunt Carrie is 89 cM segment on Chromosome 5 which seems to be my entire paternal Chromosome 5 European segment and where I am matching people with Acadian ancestry.
    My Aunt Carrie is over 80 percent African.

  14. I can’t help but wonder whether Timber is the cause of one of my daughters not showing ancestry in Great Britain, even though a great percentage of her line on both sides is from there. Her full sister’s analysis matches mine, while hers does not at all.

  15. Pingback: The Best and Worst of 2015 – Genetic Genealogy Year in Review | DNAeXplained – Genetic Genealogy

  16. Pingback: Demystifying Ancestry’s Relationship Predictions Inspires New Relationship Estimator Tool | DNAeXplained – Genetic Genealogy

  17. Pingback: Concepts – Identical by…Descent, State, Population and Chance | DNAeXplained – Genetic Genealogy

  18. This is very interesting! I have an “Unknown First Cousin” match on Ancestry (one generation older than I, matching 653cM across 35 segments), and we both have well-documented family trees covering the last several generations. There are no common ancestors on our trees, which causes me to believe there was a non-paternal event within the last couple of generations. We can rule out half of each of our trees based on not matching up with other DNA-tested first-degree relatives of one another’s, and circumstantial evidence so far is suggesting that this lady might be my half-aunt (that her father could be my grandfather — her father and my grandmother appear to have worked at the same company around the time my mother was conceived), but I’ve been troubled by the fact that we’re at the lower end of the range for that relationship, according to the Shared cM Project (although still well inside it). I also only know this lady a little bit, through some messages back and forth on Ancestry, and I think she might be wary of transferring her DNA raw data to GEDmatch for me to play with — at least not until we know each other a little better. Knowing that Ancestry’s numbers might be a little low gives me confidence that I might be right about our relationship.

  19. Pingback: Ancestry V1 vs V2 – Shared DNA and Relationship Predictions | DNAeXplained – Genetic Genealogy

  20. Pingback: Concepts – Segment Size, Legitimate and False Matches | DNAeXplained – Genetic Genealogy

  21. I am adopted and recently had my dna tested. My highest match are a family that I have 370 cm matches with and their other brother is 450 cms. would they all be 1st or 2nd cousins? thank you. Does anyone have any suggestions on finding family with little knowledge?

  22. This feature /tool isn’t showing up on my matches. There is no ‘i’ to click and see the centimorgans etc. My daughters have this tool on their matches page. I’ve tried refreshing the page.

    Why is this?

    • A segment is an arbitrary division of the entire chromosome. So there is no set number of segments. It’s like asking how many pieces are in a pie. It depends on how you cut it.

      • Roberta i just got some results back on Ancestry and the young man has 981 CM across 42 segments. I believe that he is my half nephew. I can,l with a degree of assurance rule out being his 1st cousin, he is the age of my youngest daughter. I don’t think that i am his great aunt. I believe that his father is my half brother. Any advice on how to say this to strangers from a total stranger who is this closely related by DNA?

      • I would tread very carefully and kind of test the waters first. My first question would be if he is an adoptee. Check for a tree. If none, then he might he.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s