How does 23andMe calculate the probability of illness

DNA test: 23andMe sends customers new report on their diabetes risk

So-called polygenetic risk assessments are currently a major topic in research into the relationship between the genetic makeup of people and disease risks. Not only individual genes are taken into account, but a large number of them, which together allow a prediction of the probability of certain traits, including the tendency to disease. The genetic testing company 23andMe now wants to send millions of customers a report on their risk of developing type 2 diabets, reports Technology Review online in "DNA Test for Type 2 Diabetes".

According to 23andMe, information from 1244 different locations in a person's genome is examined for the new test, each of which has a small share in the overall risk assessment. Around 80 percent of customers will find out that their specific DNA poses an average risk for them, and 20 percent have an increased likelihood of diabetes. Only the people in the group with higher risk will receive the more accurate calculation result (for example a probability of 3 to 5 seen over the entire life).

Experts express criticism of this type of report. Polygenetic assessments, while promising, are not very accurate and have no proven health benefits. "I think it's a huge experiment," says Peter Kraft, an epidemiologist at Harvard University's School of Public Health. "It's offered to millions of people, but there's still a lot we don't know." In 2013, the US government forced 23andMe to withdraw similar tests, but research has advanced and regulation has been relaxed since then.

More about this at Technology Review online:

(sma)

Read comments (33) Go to homepage
Ad ad