Just exactly how accurate are direct-to-consumer genetic tests? Francis Collins, the former head of the National Human Genome Research Institute, decided to find out for himself while researching a new book on personalized medicine. (Collins published the book The Language of God: A Scientist Presents Evidence for Belief in 2006.)
Collins, who played a central role in the Human Genome Project and is rumored to be the next head of the National Institutes of Health, announced at the Consumer Genetics Conference in Boston last week that he had had his genome analyzed by the big three of direct-to-consumer genetic testing: 23andMe, Navigenics, and DecodeMe. He ordered the tests under a fake name, lest the genomics superstar get special treatment. His speech at the conference was the first time the companies heard that they had had Collins’s DNA in hand.
Collins said that sequence-wise, the tests “appear to be highly accurate”: there were almost no differences in the genotype information generated in the three different analyses. But there were significant differences in the numbers of genetic variations used to calculate disease risk, as well as the final risk score. For example, one company used 5 single nucleotide polymorphisms, or SNPs, to calculate risk for a particular disease, pronouncing Collins at low risk. Another used 10 SNPs, placing him at high risk, and the third used 15, concluding that he is at average risk. Collins also said that the analyses provided little information on his “carrier status,” meaning whether he carried genetic risk factors that didn’t influence his own risk of disease but could be passed down to future generations.
Collins’s speech last week was more upbeat than one he gave at the Personal Genomes Conference in Cold Spring Harbor last fall. In that lecture, he emphasized the potential difficulty in finding the as yet remaining genetic variations underlying the heritability of disease. In the latest speech, he instead emphasized what could come out of genome-wide association studies: new targets for drugs.
“We have undervalued these studies,” he said at the Consumer Genetics Conference. “Even if a variant has a small impact on disease risk, that doesn’t mean it’s not a good risk target.” In type 2 diabetes, for example, two of the nine common genetic variants that have so far been linked to the disease are involved in the pathway targeted by two major diabetes drugs. “[Pharma companies and others] have not jumped on this as rigorously as they could,” said Collins. “Perhaps because it’s a bit overwhelming–there are so many of them.”
The dark secret behind those cute AI-generated animal images
Google Brain has revealed its own image-making AI, called Imagen. But don't expect to see anything that isn't wholesome.
The hype around DeepMind’s new AI model misses what’s actually cool about it
Some worry that the chatter about these tools is doing the whole field a disservice.
The walls are closing in on Clearview AI
The controversial face recognition company was just fined $10 million for scraping UK faces from the web. That might not be the end of it.
This horse-riding astronaut is a milestone in AI’s journey to make sense of the world
OpenAI’s latest picture-making AI is amazing—but raises questions about what we mean by intelligence.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.