When I finally managed to convince my sister to take a DNA test, she called me in a panic as soon as her results came back: “Yisela, I don’t think we’re biological sisters!“. I was quite surprised, and honestly intrigued. People had always commented on how similar we look to each other; we have practically the same features. She insisted I checked the website immediately, so I stopped everything I was doing and I rushed to MyHeritage.
I have to say, even after a dozen years studying DNA I was also quite surprised by how different our results were. The numbers were correct (35.0% shared DNA, 43 segments and 2,484.5 cM in common), but our ethnicity estimates showed a very different story. While my sister is 36% Greek, while I… well, I have zero Greek DNA.
In this article I’ll explain how and why this sort of thing is not unusual when comparing siblings (and other relatives’) DNA.
How DNA Inheritance Works
The DNA you inherit from your ancestors is unique to you. You receive 50% from your mother and 50% from your father, who in turn had received 50% from each of their parents, and so on.
However, genomes crossover and recombine in a random way, so members of the same family tend to have similar yet not identical characteristics. In the case of my sister, she inherited DNA segments associated with a “Greek” ethnicity that me and my brother didn’t. And the two of us have a much larger quantity of “Italian” DNA, while we three seem to share some mysterious English or Irish ancestor in 15 to 20% of our DNA.
These differences make sense. A person receives half of their markers from each parent, about a quarter from each grandparent; about an eighth from each great grandparent, etc. This mixture is different for every case because a random crossover occurs every time. That’s also the reason why, even though we are biological siblings, I only share 35% and 38% of my DNA with my sister and brother.
For you to be an exact 50% DNA match with a sibling, the same section of DNA would have to be passed down to both of you. And the furthest back in time you go, there are less and less chances of inheriting markers that correctly represent ethnicity regions distribution.
The following is another example of a person’s DNA and the DNA of their three siblings. Notice how the Europe West ethnicity region shows very different results for each person. In fact, it’s not even represented at all in the second and last results, while the sibling on the far left shows as astounding 32%:
The chart below (courtesy of Ancestry) helps illustrate how different segments of DNA might have been passed down from your grandparents to make your unique DNA.
In order to understand this example, assume each letter represents a segment of DNA and remember that:
- Not all letters get passed down to eah generation.
- The ones that do, do so randomly.
- Just because a child doesn’t have a letter doesn’t mean that an earlier ancestor didn’t have that letter.
- Siblings can have different combinations of letters.
Having more children increases the chances of passing on more of your DNA. But if you look at the graphic above, even with three children, not all of EDWARD and ANGELA’s DNA segments made it to the next generation. This is what happened with my Greek DNA. If my sister hadn’t taken the test, I would have never found out some of the relatives had migrated from that area.
We know we’re related because we have a large amount of shared DNA (most companies show you how many cMs you share with others in the database and across how many segments).
Each company’s system will try to calculate the relationship between two individuals, however, due to the random nature of DNA inheritance, relationship estimates (especially for distant relatives) are only approximate.
This is why it’s important (and more interesting) to get others in your family tested. The combination of different results can help you get a better overview of your genetic composition and your ancestors’ journey.
If you are interested in purchasing a DNA kit, we recommend you do so at the Ancesty, MyHeritage or FamilyTreeDNA, because they have the largest databases and they allow your raw data to be transferred to other sites for further analysis.