April 02, 2019
NEWBURGH, N.Y. -
Evan Merkhofer, assistant professor of Biology at Mount Saint Mary College, presents “Direct-to-Consumer DNA Tests: Hype or Hope?” at the college on Thursday, March 7.
AncestryDNA, 23andMe, FamilyFinder – the popularity of these direct-to-consumer (DTC) DNA tests has exploded over the last few years, but how accurate are the results?
Evan Merkhofer, assistant professor of Biology at Mount Saint Mary College, recently discussed the popularity and pitfalls of this new industry in his presentation “Direct-to-Consumer DNA Tests: Hype or Hope?” at the college.
According to Merkhofer, who teaches courses on genetics at the Mount, more people took genetic ancestry tests in 2017 than every previous year combined, and it is estimated that about 12 million people had taken at least one of these tests by the end of 2018. The market value for these tests by 2020 is expected to be $340 million.
“Are these tests truly a breakthrough in the fields of ancestry and medicine?” asked Merkhofer. “Or is it all just hype?”
The human genome has more than three billion base pairs, Merkhofer explained: “You’ve got a lot of DNA, so there’s a lot to look at here.” However, he added, “About 99.9 percent of one’s genome is identical to everyone else’s.”
Consumer DNA tests only look at a fraction of the three billion base pairs of the human genome – the ones most likely to bear differences. The most comprehensive of these consumer services checks about one million of these pairs.
Investigating the human genome has two major applications: Constructing genealogical linage and predicting medical issues and disease susceptibility. So in addition to offering a breakdown of one’s linage (often by percentages), when a company like 23andMe examines a person’s DNA, they will often calculate that individual’s probability of developing specific genetic disorders or illnesses, such as Parkinson’s disease.
“You’re getting a risk probability based upon how your genome sequence compares to the people they measured in a different study,” Merkhofer said. These reference groups vary by study, with people of European decent often overrepresented in these samples. With different criteria in play at each DNA test company, Merkhofer began to wonder: Just how accurate are these tests?
To find out, Merkhofer ran an experiment using his twin sisters, Lauren and Sharon. As identical twins, their DNA should be nearly the same, he said. (Small changes in identical twins’ DNA can appear over time.) And while their reports from a popular DNA test company were similar, there were some inconsistencies as well. For example, while the tests stated the sisters’ primary ethnic background was British, it said the second most prominent ethnicity was French for one sister, and Norwegian for the other.
“This is a big issue when it comes to ancestry tests like this,” he said. “Obviously, my sisters and I have the same ancestors – the exact same parents, the exact same grandparents. According to this test though, you wouldn’t necessarily see this.”
Merkhofer also gave the example of a journalist who sent her DNA to three different DTC tests. Her ethnicity was listed as 32 percent Scandinavian in one test, with another saying she was only 3.1 percent Scandinavian. The third made no mention of Scandinavia at all.
So what’s the deal? How can the same DNA produce wildly different results?
“You’ve got to keep in mind that these tests are all different,” Merkhofer said. “Different companies use different algorithms. The math is different. And they have different reference populations.”
So these DNA results are presumably accurate based on each company’s data, but with different data sets in use at different companies, results will vary. In other words, according to Merkhofer, if you’re thinking about using one of these tests, take the results with a grain of salt.
The talk was part of the college’s Investigating Research on Campus (iROC) series. The goal of iROC is to “provide a forum for Mount faculty, staff, and students to showcase their research endeavors with both Mount Saint Mary College and the local community in a manner easily understood by attendees,” explained series coordinators Merkhofer and Jennifer Park, assistant librarian for Access and Outreach services. Presentations include research proposals, initial data collection, and completed research projects.