First, Ancestry gave us centimorgans and segments. Now, they’re showing unweighted shared DNA and the longest segment. What does it all mean, and how can we use this information in our research?
This article is an illustrated guide to these genetic genealogy concepts in clear and understandable terms.
How to Find Centimorgans and Segments on Ancestry
The most basic explanation of centimorgans and segments is that they describe how much DNA you share with your genetic matches. Before we dive into terminology, let’s look at how Ancestry displays this information.
The DNA match list is ordered by shared DNA. You can see the number of centimorgans and segments at a glance. Here’s my fourth Cousin Jack, with whom I share 20 cM on 1 segment:
The same information is displayed when you open the match profile.
But things get a little more interesting when you realize that the numbers in both places are a clickable link. Judging by comments on genealogy forums, this isn’t obvious to many experienced Ancestry members.
When you click the numbers, Ancestry has long displayed a list of possible relationships that this amount of shared DNA may represent. But let’s focus on some recent additions to this pop-up display.
How to Find the Longest Segment on Ancestry
You find the longest segment on Ancestry clicking on the numbers link shown in the previous image.
Underneath the shared DNA, we now see “Unweighted shared DNA” and “Longest segment”.
I’m showing a simpler example where 1 segment is involved, and these two extra measurements are the same. That’s not always the case. But what’s also interesting here is that these numbers of 29 cM are larger than the shared DNA of 20 cM.
How can that happen? Before we get into that, I’m going to explain exactly what centimorgans and segments represent. Use the table of contents to skip these sections if you are familiar with these terms.
What are Centimorgans and Segments?
Let’s step away from the Ancestry display for a moment, and deal with the underlying genetic science.
Cousin Jack and I have inherited the same piece of DNA from a common ancestor. The more DNA we share, the closer our relationship. We measure amounts of shared DNA in centimorgans.
You have about 6,800 total centimorgans of DNA, with about half coming from each parent. So, you share about 3,400 cMs (centimorgans) with each parent and similar amounts with full siblings. The numbers are never exact, due to the random nature of inheritance.
My fourth Cousin Jack and I share a rather small amount of DNA, which we can map out on a specific chromosome. Our shared DNA is nestled on chromosome #12, which is one of our 23 pairs of chromosomes. I’ll break it all down using this schematic diagram:
The diagram chops up the chromosome into minute positions and shows (in purple) where our shared DNA starts and ends.
A Simple Explanation of Segments
The start and end position of shared DNA defines a single segment.
So, there you have the meaning of a segment. I only share one segment with Jack.
In an alternative scenario, we could also have shared DNA elsewhere on the chromosome. Or we might have a shared piece on a completely different chromosome. The start and end positions of shared DNA marks out those additional segments.
A Simple Explanation of Centimorgans
Segments will vary in length, and we want to know how long these segments are. The diagram chops up the shown segment into 20 genetic units. Those are the 20 shared centimorgans between myself and Cousin Jack. Simple as that!
Centimorgans aren’t really a measure of distance or length. You’ll get along fine if you think about them as a consistent unit of shared DNA, where Cousin Jack at 20 cM is a more distant relation than Brother John at 3,460 cM. If you want to get more technical, check out the definition from the International Society of Genetic Genealogy.
A Video Guide to the Ancestry Longest Segment
Here is a companion video to this article. The rest of this article covers the topic in more depth.
What is the Ancestry Longest Segment?
The longest segment is quite a simple concept, but the display on Ancestry may be a little unclear at first. Let’s look at the simplest explanation first. We’ll come back to Cousin Jack later, but we’ll switch focus now to my more distant Cousin Virginia.
This is a real example. I’ve chosen Virginia because she has also tested with MyHeritage. Unlike Ancestry, MyHeritage has a chromosome browser which shows me which chromosomes have our shared DNA. That lets me give you a more accurate diagram.
Here is Cousin Virginia on Ancestry:
We share 19 centimorgans across 2 segments, with the longest segment being 11 cM.
The MyHeritage numbers are very similar (this is not always the case). More specifically, MyHeritage tells me that our shared segments are on Chromosome #1 and #17. The larger segment is on #17.
Definition of the Longest Segment
The longest segment is simply the shared segment with the highest number of centimorgans. Ancestry displays the longest segment as 11 cM. Virginia and I only share two segments, so I can infer the approximate size of the other segment.
What about the other segments?
The Ancestry display doesn’t let me infer other segment sizes when there are more than two involved. Take my closest relative on Ancestry:
Ancestry tells me that our longest shared segment is 45 cM, but I have no idea how the rest of our shared DNA is spread across the other 19 segments. That’s because Ancestry does not provide a Chromosome Browser. There are ways around this – see my article on four ways to use a chromosome browser with your Ancestry DNA matches.
You may not have forgotten the elephant in the room, my Cousin Jack (sorry, Jack). We share a total of 20 cM across 1 segment, but the longest segment is reported as 29 cM. What the heck? Before we get to that, we need to look at that other figure in the Ancestry display: “Unweighted shared DNA”.
Timber and Unweighted Shared DNA
There are several reasons you may share a piece of DNA with a genetic match in the Ancestry database.
What you want for genealogy research is that the DNA is identical by descent (IBD). This means that you inherited the DNA from a common ancestor.
There is also a possibility that the DNA is identical by chance (IBC). The smaller the segment, the more likely this will occur. This is what leads to false positives in your low cM matches i.e. people with who you share DNA with but who are not genealogically related to you at all.
But we won’t worry about that thorny issue in this article. There’s a third type of DNA sharing we need to address.
Identical by Population
That third type is IBP, or identical by population. This refers to a segment of DNA that is common across a particular ethnicity or population group. Yes, there is a common ancestor. Indeed, there may be many common ancestors in different lines. But these ancestors are so far back in time, that you will not be able to evaluate or identify them using DNA matches.
For a detailed explanation, go check out this article from the DNA Explained blog.
Ancestry’s Solution: Timber!!
Ancestry’s DNA processing attempts to filter out shared DNA that is identical by population. Their algorithm is called Timber.
Let’s take another look at Cousin Jack and his strange numbers.
Ancestry DNA processing found 29 cM of shared DNA between us. Then Timber goes to work.
The Timber algorithm trawls through the Ancestry database to see who else shares those centimorgans on that particular chromosome pair. The algorithm has decided that a fairly sizeable 9 cM is prevalent across a particular group of people and is due to some very distant ancestor.
A quick look at the ethnicity tab tells me that Jack and I only share the Ireland region of Ancestry’s ethnicity groupings.
I assume that Timber is slicing out a segment of DNA that is prevalent amongst many other DNA testers with Irish heritage.
Definition of Unweighted Shared DNA
The unweighted shared DNA is therefore the number of shared centimorgans before the Timber algorithm is applied.
How Can the Longest Segment Be Bigger than the Total Shared DNA?
It’s also clear that the Longest Segment is showing the unweighted number. This is why it can be bigger than the total shared DNA.
I’m not sure that Ancestry will continue with this display. It’s just confusing. They should either change the labels or show more numbers i.e. all the unweighted and weighted numbers.
How Useful is the Longest Segment?
Now that you know what it is, the question becomes: do you need to know? In other words, is this information useful when researching your DNA matches?
Not necessarily. If you do not have significant levels of endogamy in your ancestry, then you can gloss over this measure. Keep focusing on the shared centimorgans in the main display.
However, it becomes far more useful when your research must account for endogamy across your lines.
Using the Longest Segment with Endogamy
Endogamy refers to significant levels of inter-marriage within a group for many generations. This may be due to reasons such as an isolated location, religious traditions or cultural norms. On a practical level, it may become evident to you if you identify multiple marriages across different branches of your family tree.
Even if you can’t decipher prior generations, it may become apparent if you share “too much” DNA with known relatives. This will happen if you are related to, for example, a second cousin through multiple branches on both your paternal and maternal lines.
Some groups and communities have a growing resource of research and articles to help genealogy research with endogamous heritage. You’ll find many excellent articles that focus on Jewish or Acadian genealogy.
I’ll also give a shout-out to this detailed article on Polynesian heritage. There are lots of screenshots from Ancestry, as the researcher uses the new longest segment size to analyze DNA matches.
The guideline is to use the longest segment to zone in on DNA matches that are worth researching for a recent common ancestor. The longer the segment, the more likely that you share a more recent ancestor with the DNA match. Well, that’s a little vague. How long is long enough?
There’s no hard and fast rule, but a figure of 20 cM is often mentioned by genealogists who research endogamous trees.
When you click on any centimorgans link, the pop-up window gives you a breakdown of the possible relationships at that number of centimorgans.
The display is laid out in descending order of probability. Here’s the top two for myself and Jack.
Judging by the questions on genealogy forums, Ancestry testers don’t find this display very helpful. I agree. I can’t easily visualize where a “half 2nd cousin 2x removed” would fit into my tree. So, this is not my first port-of-call when I’m trying to figure out a relationship.
My Favorite Online Centimorgan Chart – Shared cM Project
I like charts that arrange centimorgans in the style of a family tree. My go-to online chart is the Shared cM Project on the DNA Painter website. The chart is interactive (roll your pointer over it). And when you enter a cM total into the input box, it highlights the possible relationships in the tree presentation.
My Favorite Offline Centimorgan Chart – the Green Chart
If you want a chart to use offline or to print out, then you can’t go wrong with the “green” version from the DNA Detectives Facebook group. Here’s a link to a copy outside of Facebook. The DNA Geek blogger has permission to host the image.
Differences Between Charts
Different charts use different estimates for relationships. These differences become apparent in the upper and lower ranges of the estimates of how many centimorgans may represent a relationship.
Blaine Bettinger’s Shared CM project (mentioned earlier) is compiled from user-submitted data and is more likely to cover “unusual” edge cases when compared to the chart from DNA Detectives. And Ancestry’s predictions are partly proprietary.
Just be aware that the edges between ranges will overlap, and these charts provide a frame for your genealogical research.
Centimorgans And Percentages
Sooner or later, you’ll realize that shared DNA can also be reported in terms of percentage. For example, 23andMe displays percentages for your DNA matches in their database. If you’re used to one format, it can be difficult switching mentally to the other.
The close relationships are clear enough. You share about 50% of your DNA with a sibling and about 25% with a grandparent.
Personally, I much prefer to use centimorgans, because I have very few close DNA matches on either Ancestry or 23andMe. When you’re down in the weeds, the percentages drop below 1%. I find these more difficult to distinguish between one and the next.
How to Convert Centimorgans to Percentages
It’s easy to do an approximate conversion between the two representations. I mentioned that we all have about 6,800 centimorgans in our entire DNA.
To convert any figure of centimorgans into a percentage, divide the cM by 6,800 * 100. Or just divide by 68 for less typing into that calculator.
You share 3,450 cM with Brother John? That’s 3450/6800*100 = 50.7%.
Cousin Jack and I share 20 cM, which is 20/68 = 0.29%.
How to Convert Percentages to Centimorgans
MyHeritage, bless them, show both percentage and centimorgans in their match display. But 23andMe show percentages in their shared match display.
To convert the percentage to the approximate number of centimorgans, just multiply by 68.
Quick Approximate “How Many” Answers
These are averages without the ranges. Use them for a quick check, but be sure to evaluate the upper and lower levels given by the more complex charts in the prior section.
|How many centimorgans shared with a parent?||50%||3,400|
|How many centimorgans shared by siblings?||50%||1,700|
|How many centimorgans shared between half-siblings?||25%||1,700|
|How many centimorgans shared with an aunt or uncle?||25%||1,700|
|How many centimorgans shared with grandparents?||25%||1,700|
|How many centimorgans shared with a first cousin?||12.5%||850|
|How many centimorgans shared with a first cousin once removed?||6.25%||425|
|How many centimorgans shared with a second cousin?||3.13%||213|
|How many centimorgans shared with a 2nd cousin once removed?||1.5%||102|
|How many centimorgans shared with a third cousin?||0.78%||53|
|How many centimorgans shared with a fourth cousin?||0.2%||14|
16 thoughts on “Ancestry Centimorgans and Segments (and Longest Segments)”
Thank you! But what about when longest segments is higher cM’s than total or maybe fills up total close to top and ancestry has identified several eg more than one and I’ve seen up to four segments and am left totally confused wondering where on earth Ancestry has fitted in the other 2 or 3 segments with no, or only like 3 cM of space left in total outside of longest identified?? And where total number of segments hasn’t changed, but now can see unweighted which I’ve seen up to twice as big in cM as total? And in some I’ve seen no change in total or number of segments and only difference is can now see length of largest (total and weighted exactly the same) and appears room for other segments to fit? Will appreciate if you can explain what each of these scenario’s mean.
My sister matches Lance (5th cousin) at 10 cM over 1 seg and matches his mother (4th cousin 1x) at 7 cM over 1 seg. Both have Unweighted DNA and Longest Segment of 10 cM. I don’t usually even look at Unweighted and Longest Segment since I don’t know what to do with them – as mentioned they are confusing. In this case, the Unweighted DNA measure explains this anomaly where a son scores better than a mother.
It seems like ancestry’s Timber filter algorithm may be causing some confusion. Why did it eliminated 3 cM of Lance’s mother but did not eliminate the same DNA for Lance?
This article was written before ancestry decided to add in percentage of shared DNA – which is duplicated by the cM divided by about 69. Now 99.885% of my DNA matches show that I share ‘< 1%' of DNA with matches. Makes me not interested in doing DNA genealogy if that many matches share so little DNA with me.
Statistically, you will always have far more DNA matches at the lower range of cMs. I find that people with 30K matches or more tend not to look at the lower end. But some of us (like me) have a much smaller number of matches overall, so we spend more time down there.
I find that the trick is to quickly check if the match has a large linked tree. The larger trees may go back enough generations to let you investigate the links.
I’ve managed to build my own tree back several generations thanks to the excellent research of a match of 8 cM (after I’d verified all his documents). But I may have examined hundreds of other low CM matches without success before I hit gold. Everyone will decide on whether this is a good use of their time!
If two individuals share a longest segment of 109 cMs what is the most likely relationship that they would share? Could this indicate a half aunt/half niece relationship? And if so would the half aunt be the daughter of the half niece’s grandfather? The answer to these questions will hopefully solve the question of my mother’s paternity. I am convinced that the grandfather of this half niece is my mother’s father. Please shed some light on this for me. I would be grateful if you could help solve this riddle. Thanks.
109cm is quite low, and seems too low to be half aunt/half niece which tends to have a minimum of 500cm according to DNA Painter.
When I plug 109 into DNA Painter to see the possibilities, there are 22 relationships that could fall into that level of cM. Shared matches may help you narrow down the list, but you’ll need to continue with standard genealogy research to identify the relationship.
I have just received an email saying I have a shared 1833cm with her 60 segments ,the longest being 104cm long.
Could you say how close a relation this would be.
That is a close relationship. I like to use the DNA Painter Shared CM tool here.
You just plug in the number of centimorgans and it gives you the statistical probabilities.
It lists these relationships as most likely: Grandparent / Aunt / Uncle / Half Sibling / Niece / Nephew / Grandchild.
Hope that helps. Take a deep breath and take care of yourself. I have an article on DNA tests for adopted adults with some advice on surprise findings. Some of it may be relevant.
Hi: How easy is it to find great-great-grandparents using Ancestry DNA results? I have speculated for years who they were and finally a person showed a DNA match of 17cm/1 segment that went through them. His third-great-grandparents would be my 2nd-great-grandparents. This seams consistent. To what degree of certainty can I claim them in my family tree?
Researching your great-greats will require building a family tree with the help of your DNA results. There’ll be a lot of genealogy research i.e. looking up census records, who married whom, shipping passenger lists etc.
How much help your DNA results will give you largely depends on your heritage. Let’s look at two extremes.
It will be much easier if your great-greats were born in the United States or their children were. You will have many thousands of DNA matches on Ancestry.com and some of them will have well-researched family trees going back generations.
It will be very difficult if your parents are first generation Americans and their heritage is from countries where there isn’t a strong history of civil administration records (birth, marriage, death).
I can’t get my head around segments etc. so could you please tell me what the following means and what probable relationship I would have with this person. There is a range of possibilities given when I click on this.
2% shared DNA | 171 cM across 10 segments
Unweighted shared DNA: 171 cM
Longest segment: 38 cM
You don’t need to pay too much attention to the segments, focus on the centimorgans. There’s no way to pick the most probable relationship on DNA alone. I assume the person doesn’t have a public tree for you to examine? You can usually rule out a few possibilities if you can glean the person’s age.
On my maternal grandfather’s side, from the Scotch Irish population of the southern colonial United States, we have hundreds of matches who share only 1 segment with us, which is over 50 cm. I have found it up to 92 cm on its own. Just to confirm, there are no other shared segments, just the one big guy. I have this segment in my matches, as do my mother and her brother.
Many articles say that the larger the segment, the closer the relationship, but I am so puzzled because the trees do not have common ancestors. Most of them don’t even share surnames, just the same areas of Virginia, The Carolinas, and Missouri before 1850.
Do you have any experience with one giant, sticky segment? I feel that this could just be one single piece of endogamous dna that is being passed along, but I don’t know if that is possible due to the sheer size of it.
Thank you so much for sharing your wisdom; we have endogamy on both sides of our family and your articles have helped guide us immensely.
I don’t want to speculate because I haven’t seen something like this. Very interesting though.
I am new to DNA. I have a match of 1,510 cM across 38 segments 1,510 cM unweighted shared DNA and 179 Longest segment. What does this mean? How is this person related to me? Thanks for any help on this.
That is quite a close match. I recommend that you input the number into the Shared CM tool at this link.
When I do so, the tool says the relationship is highly likely to be one of any of these:
You may be able to rule some out due to your age.
I wish you the best of luck in your next steps.
Because this relationship may be a surprise to your DNA match, I suggest that you take screenshots and as many notes as you can before reaching out e.g. via an Ancestry message.