Do you have DNA results from multiple companies and noticed differences between them? In this post we will explain why.
People who are passionate about DNA testing often conduct multiple DNA tests with different companies or upload our DNA data from one provider to another, in order to maximize results and ethnicity estimates. Doing this allows more DNA matches and ethnicity estimates.
However, we quickly become aware that ethnicity estimates from different companies can differ dramatically – sometimes we even see entire regions missing from one set of results!
Although we understand it shouldn’t be expected that our ethnicity estimate would exactly mirror our family tree, it can still be frustrating when things don’t seem to match across companies. How could results from different websites vary when our DNA remains the same?
Note that this information only refers to our ethnicity or ancestry estimation and not DNA matching. Each company calculates their DNA matches similarly, so shared DNA should not differ greatly between providers.
Why Your DNA Results Differ Between Companies
There may be various reasons for why Ancestry DNA results won’t match those from 23andMe or Family Tree DNA, for instance. Below you will find these various explanations listed and described further in more depth.
Use the information presented below to reexamine your DNA results from each company, and determine which are most applicable to your results. This could help shed new light on your family’s DNA story.
Every company utilizes different populations as samples in order to calculate the estimated ethnicities of its customers. Each company uses its own method for selecting whom it includes in its sample population and has various resources for gathering samples.
DNA testing companies provide us with an ethnicity or ancestry estimate, which can tell us where our ancestors may have come from. Although the exact process varies from company to company, the basic idea behind testing our DNA involves comparing it against people living elsewhere around the globe so as to discover where our ancestors may have come from.
With different inputs – sample populations from which to draw results – coming together, different results for ethnicity analysis will surely emerge.
Some sample populations are larger than others Some of the larger DNA testing companies, like Ancestry DNA and 23andMe, can include many thousands of DNA samples for our estimates to calculate. Their respective websites indicate this to be true as of 2022: AncestryDNA uses over 68,000 and 23andMe utilizes 14,000+ samples.
These numbers only refer to the general regions we see on our results. Some companies, like Ancestry and MyHeritage, also include features which assign us Genetic Communities or Groups based on samples included in their customer database – these features may require much larger calculations due to such extensive customer databases.
Typically speaking, ethnicity estimates generated with larger sample populations are more accurate.
Each company has different criteria for selecting its sample populations; all DNA testing companies adhere to an official policy that only includes samples from people living in regions around the globe who share ancestry from that location. As we don’t know exactly how these samples are selected by each company, variations in results could be attributable to differences between companies’ inclusion requirements for inclusion into sample populations.
As you have learned in previous sections, each company uses its own reference panel. Our DNA is compared against that of an entire population to estimate our ethnicity estimate; this process does not directly involve comparison between individual DNA samples and us to produce an ethnicity estimate.
Instead, most companies utilize something known as a “reference panel”, consisting of DNA from various sample populations to compare to our DNA samples. If any part of ours matches those from certain areas in the reference panel, that percentage will show up on our results as “matches from that region”.
Technology used by each company makes things even more complex. All companies strive to achieve accurate, precise results, so they work on developing methods for estimating ancestry that offer maximum precision and accuracy.
Technology used for creating reference panels varies across companies; while their overall process may be similar, each has their own proprietary method for creating reference panels to assess ethnicity and ancestry. Some of them offer very transparent processes with published papers describing it in simple language so anyone can easily comprehend.
Read up on Ancestry’s ethnicity estimation process in their Ethnicity White Paper and 23andMe’s in their White Paper about their Ancestry Composition algorithm (link opens PDF download).
Every DNA testing company labels regions of the world differently, often due to the different names or definitions given for different regions by various companies. We may see multiple areas combined into one region on one DNA testing website while separated out as individual regions that roughly align with modern country borders on another one.
Differences in terminology for similar regions
23andMe’s current definition for France & German includes both France & The Netherlands while Ancestry DNA places Germany and The Netherlands into “Germanic Europe.”
But on Family Tree DNA, part of the Netherlands appears on its map under England, Wales, and Scotland while much of Germany falls under “Central Europe.”
One with German and Dutch ancestry may mistakenly interpret results from these three companies as different; however, these regions define differently but should yield similar outcomes for anyone with these backgrounds.
Some testing companies don’t report small percentages Some companies have adopted policies not to report small percentages, such as less than one percent, that correspond to specific regions – often called trace regions – when performing testing services.
Very small percentages could indicate either small error, or they could indicate distant ancestry. Many DNA testers, myself included, are fascinated with testing these areas as it could reveal something about our ancestors that we wouldn’t otherwise discover.
23andMe and MyHeritage are examples of DNA testing companies that offer results with small percentage matches; MyHeritage displays regions matching at less than 1% while 23andMe offers matches down to as low as.2%.
Conclusion
I hope that this post has provided you with enough knowledge to comprehend why DNA results from different companies vary so much, yet they all seem similar. You now understand it’s perfectly normal and doesn’t indicate whether one company’s results are more or less accurate or inaccurate.
If you have any inquiries or observations to share about something you read here, or would like to share something specific that worked well in your situation based on what I mentioned here, I welcome your participation in the discussion below.
Leave a Reply