If you could find out if you were going to develop a fatal disease, would you want to know? Think about it. Or would you require your fiancé to have a DNA test before you would agree to marry him?
Many of us would shun such knowledge as TMI—too much information—or such demands as impertinent. Yet that kind of genetic sleuthing has been proposed by startup company Genomic Predictions, initially to determine one’s likelihood of developing Type 1 diabetes. This DNA test, however, is not designed for use on adults who might want to learn of their own propensity to disease, but for use on human embryos created through IVF, to evaluate their suitability for implantation. As in a nightmarish college admissions scenario run amok, “doctors and parents could huddle and choose to avoid embryos with failing grades.”[1]
Powerful Computing Meets Massive DNA Data
The technology, called expanded Preimplantation Genetic Testing (ePGT), uses volumes of genetic data and powerful computer models to predict which IVF embryos will develop Type 1 diabetes. Unlike Preimplantation Genetic Diagnosis, which looks for a specific disease carried by one or both parents or for a chromosomal anomaly such as Trisomy 21, ePGT technology examines an array of genes that contribute not only to a disease, but also to characteristics such as eye color, height, or intelligence.
Creating the model for these calculations requires an enormous amount of genetic data. And where do you suppose that data comes from?
One source is the U.K., specifically Biobank, which recently released medical records and DNA information for 500,000 middle-aged British people. Included in the data was information about SNPs, or single-nucleotide polymorphisms, small DNA mutations that differ slightly from person to person. SNPs are thought to be linked to matters as serious as life-threatening diseases and as trivial as bread consumption.[2]
The other DNA treasure trove? Data from consumer do-it-yourself DNA test kits, sold by companies such as Orig3n, 23andMe, and Ancestry.com. These companies do not operate simply to sell consumer DNA kits so that companies can target purchsers with ads, but also to sell data to medical research firms. As one writer describes it, 23andMe’s “big moneymaker” is salable consumer data: a company spokeswoman said that 80 percent of its two-million-plus customers consented to having their data used for research.[3] In 2015, the company sold customers’ genetic information to Genentech for $10 million.[4]
The mad rush to spit into a test tube shows no signs of slowing down. Ancestry.com sold over 1.5 million DNA test kits during the Black Friday–Cyber Monday weekend last year, and 23andMe’s kit is a top seller on Amazon.[5] Customers pay for DNA results, which the companies turn around and sell to research firms.[6] The DNA investigation can include not only hunts for ancestry, but also analysis of personal health data. That might explain 23andMe’s sale of data to Pfizer and Genentech, and Ancestry’s deal with Calico, a longevity project created by Alphabet, the parent of Google. If you have long-lived ancestors, or ones with remarkably short lives, your DNA is particularly desirable.
Illusory Privacy of DNA Results
Even if a customer opts out of every possible use of his or her DNA, most companies retain the right to use or sell it as “anonymized” data, with your name and other personal information removed from the sample. Yet a determined researcher or biohacker could still discover your identity. Several years ago, a group of researchers was able to determine a person’s last name from a DNA sample and then triangulate for his identity using only publicly available information.[7] Even the most vigilant customer, one who reads all the fine print and opts out of every option, is unwittingly contributing to the most massive accumulation of genetic information in history if he sends his sample to a company that retains the right to use anonymized data.
How much do people actually think about what they are signing up for? In 2015, over 90 percent of customers consented to their data being used for commercial and non-profit research purposes.[8] That number has since declined, so now a startup has announced plans to facilitate customers’ selling their genetic info themselves, using cryptocurrency, something like bitcoin.[9]
Genetic Profiling Without Limits
When companies like Genomic Predictions combine access to an unprecedented volume of DNA data with massive computing capabilities, the results can be startling. Using machine-learning techniques, Genomic Predictions physicist Stephen Hsu and bioinformatician Laurent Tellier claim to be able to predict a person’s height to within 1.5 inches, using only a DNA sample.
Height may be the easiest non-disease characteristic to predict, but Hsu is interested in testing for an even more worrisome trait: IQ, which is much harder to predict than height. Hsu says that even though his company will not select for intelligence, “I fully predict it will be possible.” The most likely early adopters of this technology when it matures? “Billionaires and Silicon Valley types,” who will “do IVF even though they don’t need IVF.”[10] The pursuit of the smartest, tallest, and healthiest children may initially come with a hefty price tag, including popular resistance. But if the rich succeed, Hsu suspects that the rest of society will fall in line.
Nor is genetic profiling limited to the testing of already-created embryos. In 2013, 23andMe was awarded a patent for genetic and computer technology that would enable parents to handpick an egg or sperm donor. An analysis of the parents' genetic info combined with that of the potential donor could indicate whether a particular disease could be avoided. Powered by massive computing, genetic calculations can match the intended parents' DNA with DNA profiles of large caches of potential egg and sperm donors, narrowing the field to a "preferred donor." The drop-down menu included in the patent application included options not only for disease traits, but also for eye color, lactose tolerance, and muscle performance.[11] Even though designer offspring is more illusion than reality, it is unsettling that a patent has been awarded on the technology that would enable parents to choose.
Paradoxically, expansion of choice may result in contraction of parents’ hospitality toward their own child. As law professor Dov Fox points out, such technology inherently undermines parental acceptance of their child. When parents select for a particular kind of child—tall, with a sense of humor and a talent for running marathons—they are less open to the “possible contours along which their child’s life may unfold.”[12] After all, once a person has conceded the morality of selecting for (or against) one trait, however serious, then the door has been opened to selecting for multiple traits—and expecting the results to match the selections.
Ironically, Genomic Predictions’ co-founder Tellier was inspired by the movie Gattaca, a dystopian sci-fi portrayal of a future society divided between the genetically perfect and the “de-generates.” In the movie, the human spirit triumphs over the pursuit of perfection. Yet, as happens with the genetically designed children in Gattaca, Tellier’s goal is to notify parents if their (embryonic) offspring are “outliers” or “at the wrong end of a statistical curve for disorders such as diabetes, schizophrenia, dwarfism, and late-life osteoporosis.”[13] I would be among the 54 million Americans with osteoporosis.[14] “Outliers,” indeed.
Setting aside the pros and cons of amassing reams of genetic information, privately or publicly, there is still the question of the ultimate uses to which such data would be put. While much of the research would aim at eradicating serious diseases, other, more dangerous kinds of research would also be done. Finding ways to extend life may help those who are at risk of premature death. But the attempt to extend a normal lifespan to abnormal lengths will not be far behind. Neither will the desire, followed closely by the obligation, to select the “best” offspring. The eugenic impulse was not only a mark of the early twentieth century. It is an indelible blemish on fallen human nature.
References
[1] Antonio Regalado, “Eugenics 2.0: We’re at the Dawn of Choosing Embryos by Health, Height and More,” Technology Review (Nov. 1, 2017): technologyreview.com/s/609204/eugenics-20-were-at-the-dawn-of-choosing-embryos-by-health-height-and-more.
[2] Regalado, “Eugenics 2.0.”
[3] Kristen V. Brown, “23andMe Is Selling Your Data, But Not How You Think,” Gizmodo (April 14, 2017): https://gizmodo.com/23andme-is-selling-your-data-but-not-how-you-think-1....
[4] Deborah C. Peel, “The Hidden Danger of Do-It-Yourself Genetic Tests, Newsweek (Dec. 16, 2017): https://newsweek.com/hidden-danger-do-it-yourself-genetic-tests-749475.
[5] Megan Molteni, “Ancestry’s Genetic Testing Kits Are Heading for Your Stocking This Year,” Wired (Dec. 1, 2017): https://wired.com/story/ancestrys-genetic-testing-kits-are-heading-for-your-stocking-this-year.
[6] It is important to note that population-wide genetic studies can yield significant findings about health that contribute to the common good, and that such findings would not otherwise be obtainable.
[7] Melissa Gymrek et al., “Identifying Personal Genomes by Surname Inference,” Science (Jan. 18, 2013): http://science.sciencemag.org/content/339/6117/321.
[8] Katie M. Palmer, “Another Personal Genetics Company Is Sharing Client Data,” Wired (July 21, 2015): https://wired.com/2015/07/another-personal-genetics-company-selling-client-data.
[9] Christina Farr and Ari Levy, “You’ll soon be able to sell your genetic info for a new cryptocurrency called Luna Coin,” CNBC (Dec. 18, 2017): https://cnbc.com/2017/12/18/luna-coin-project-sell-your-genetic-data-for-crypto-tokens.html.
[10] As quoted in Regalado, “Eugenics 2.0,” note 1.
[11] “Gamete donor selection based on genetic calculations,” U.S. Patent 8,543,339 (Sept. 24, 2013): https://google.com/patents/US8543339.
[12] Dov Fox, “23andMe’s Designer Baby Patent,” HuffPost (Oct. 4, 2013): https://huffingtonpost.com/dov-fox/23andmes-designer-baby-pa_b_4042165.html.
[13] Regalado, “Eugenics 2.0,” note 1.
[14] National Osteoporosis Foundation, “Bone Health Basics: Get the Facts”: https://nof.org/preventing-fractures/general-facts.