Is genome confidentiality such a big deal?

I recently sent my spit sample to 23andMe in order to be genotyped. I chose the option to share my genotypic information with 23andMe and all its clients (*). This brought me to think a bit about genome privacy, despite not being involved in genomics research myself.

Genome confidentiality seems to be a big deal. For instance the 1000 genome project went into great lengths to completely anonymise the data; and no phenotypic information is provided in order to preclude identification (which seriously limits scientific discoveries possible from the project’s data). It seems that showing your genome is more taboo than showing your winky. Once it is out, it will be used by plenty of oppressors etc. This fear is increased by movies such as, otherwise excellent, Gattaca. The most frequently voiced concern, because the more immediate maybe, is the fear that insurance companies would base their premiums on the genome.

This fear is probably real; insurance companies will try to base their premiums on the genotypes. However, I would argue that 1) much information about your genome is already out there, and insurance companies already use it, 2) use of genome information is perhaps not a completely outrageous idea, and 3) there are other ways than confidentiality to protect us.

Although the relationships between genotype and phenotype are complex, our entire modern understanding of biology is that our phenotype is the expression of our genotype in the environment. Accordingly one can infer much about your genome from your phenotype. You gender, the colour of your skin and hair, the fact that you are obese or not, send signals, more of less ambiguous, about your genome. Your blood type is generally fairly public (in France we carry a “blood ID card”, in case we have an accident). Plenty of other phenotypes can be swallowed by good machine learning algorithms that will infer part of your genotype. Insurance companies already use that. It is more apparent for, e.g., car insurances, for which women pay lower premiums. But each time you fill a form for an health insurance mentioning past illnesses or behaviours, or illnesses of your relatives, you are actually feeding information about your genotype. And insurance companies already use that to alter what they cover.

Now, is it really a problem? Why should private insurance not change premiums based on the genome? They are private companies, and their mission is 1) not go belly up, and 2) make money (whether we are talking of indecent amounts of money is another issue). The situation is very different for national insurances. In this case, everyone pay, proportionally to their income, and everyone is covered no matter what. And people must subscribe to national insurances (BTW, this is why French securite sociale, the NHS, Obamacare etc. are so important. Society cannot function properly without basic national egalitarian systems). But no-one is forced to subscribed to a private insurance. The insurance company then needs to balance between number of clients and premiums. If the premiums become too high because of their genomes, people at risk will not subscribe. And people who are not at risk will also not subscribe … because not at risk, so the company will go down. Companies would also compete, which would bring the price down. Some would also be advertising the fact that they do not take so much account of the genotype etc.

However, it is a controversial issue, and the society at large may decide that insurance companies must not base their premiums using genotyping info. But is genome confidentiality the solution to enforce that? There are laws forbidding the use of some data for discrimination. People cannot be discriminated based on their ethnic origin for instance, or their religion. As such, when they subscribe to a health insurance, they do not need to come to the meeting with a whole body suit. Why would it be different for the genome? If a law says “Insurance companies are not allowed to use genome derived information”, any insurance company caught doing so can be prosecuted and pay lots of money to the relevant customers. As far as I am concerned, I prefer to live in a world where the apples are on display and not locked away in the market, but if an apple is stolen, police try to catch the apple thief.

In the science fiction novel The Light of Other Days, tiny wormholes allow to see in the past. Effectively this allow anyone to spy on everyone … a fraction of second in the past. All privacy is gone. And then humankind reacts in two ways. Half of the humankind freaks out an start living naked in crowded entirely dark houses. The other half realizes that privacy if gone for good and adapts. They grow up.

(*) I did that despite a strong aversive reaction to the recent announcement of a patent deposited by the company covering “Polymorphisms associated with Parkinson’s disease”. The “discovery” (and not the “invention”) was made using the data provided by thousands of clients who paid for being genotyped. The company did not deposit the patent in order to protect the community from exclusive exploitation, but instead to ensure it guarantees a return on investment. The reason advanced is that treatments will only be developed if such a return on investment is sufficient, and not to stiffle innovation or hinder future research. Past history such as Myriad Genetics and the BRCA genes makes me somehow dubious. I will only believe that claim if 23AndMe launches a competitive process to select the company that is the best suited to develop treatment, and gives an exclusive license to this company. But 23AndMe itself should not earn money from that patent.