{"items": [{"author": "David Chudzicki", "source_link": "https://plus.google.com/106120852580068301475", "anchor": "gp-1362336079523", "service": "gp", "text": "I would imagine that soon you'll be able to get cheaper life insurance if you're willing to provide your genetic information (and ifnit doesn't have bad signals). Eventually this reaches a point where it's impossible or unaffordable to obtain life insurance without doing so. \n
\n
\nOr they regulate this opportunity away and the above isn't possible. Either way, life insurance soon isn't a reason for genetic privacy. ", "timestamp": 1362336079}, {"author": "Mad", "source_link": "https://plus.google.com/100533872344198336746", "anchor": "gp-1362337421177", "service": "gp", "text": "That's actually a serious inaccuracy in the Snyder story: his life insurance incident only occurred after he had a diagnosis of diabetes. Not due to any genetic information. Although GINA doesn't protect against genetic discrimination in life and long term care insurance, so far I haven't heard of any companies doing so.\n
\n
\nIf/when insurance companies start to discriminate, they would almost certainly be doing so for much more well-established risks -- like the ApoE4 variant (much higher risk of Alzheimer's). Snyder's diabetes prediction was the result of their mostly-blackbox internal genome interpretation methods for combining risks for multiple variants. (The more general issue of how to combine information from multiple variants is also extremely unclear.)\n
\n
\nYou might also be interested in module 6 of the PGP study guide (more \"plain English\" than the consent forms): \nhttp://www.personalgenomes.org/exam/v20120430-study-guide.pdf\n
\n
\nI think one of the most significant issues people might not realize is that our ability to interpret genome data is extremely limited right now. So you could think \"oh, nothing serious in there\" and share it, only to have research advance and later discover there was information in there that you regret sharing.\n
\n
\nOur study guide concludes with currently-sci-fi hypothetical risks: planting samples of your DNA to frame you for a crime, or using your DNA or cells to create IVF children or clones without your permission.", "timestamp": 1362337421}, {"author": "Jeff Kaufman", "source_link": "https://plus.google.com/103013777355236494008", "anchor": "gp-1362341882302", "service": "gp", "text": "@Madeleine\n\u00a0\"you could think 'oh, nothing serious in there' and share it, only to have research advance and later discover there was information in there that you regret sharing.\"\n
\n
\nI guess the question is, what sort of information could someone possibly get out of my genome that I would rather have kept secret?\n
\n
\n\"planting samples of your DNA to frame you for a crime\"\n
\n
\nI wonder if that's actually much easier for a potential framer than just getting some of your hair or skin or something.\n
\n
\n\"create IVF children or clones without your permission\"\n
\n
\nI don't think this bothers me.", "timestamp": 1362341882}, {"author": "Mad", "source_link": "https://plus.google.com/100533872344198336746", "anchor": "gp-1362342626089", "service": "gp", "text": "\"what sort of information ... would rather have kept secret?\"\n
\n
\nPeople aren't usually secretive about physical illness, but calling into question your mental state might be more distressing. i.e. risk factors for mental illness and dementia.\n
\n
\nThat said, the PGP has some ApoE4 homozygotes who have released their genetic data publicly. In the end I think many people become comfortable living with these risk factors as another part of their identity.\n
\n
\nCounterpoint to framing-for-crime: by releasing data/samples publicly you now have an alibi?\n
\n
\n\"I don't think [IVF children or clones] bothers me\"\n
\n
\nYeah, a lot of people say \"Cool! Sign me up!\" to the hypothetical risk of children or clones. :-)", "timestamp": 1362342626}, {"author": "Jeff Kaufman", "source_link": "https://plus.google.com/103013777355236494008", "anchor": "gp-1362342984278", "service": "gp", "text": "@Madeleine\n\u00a0\"by releasing data/samples publicly you now have an alibi?\"\n
\n
\nThis hasn't come up either way in court yet, right? \u00a0Partly because the cost of dna synthesis is still way high?\n
\n
\n\"a lot of people say 'Cool! Sign me up!' to the hypothetical risk of children or clones\"\n
\n
\nI'm not quite that gung ho about unknown genetic offspring, but it doesn't bother me either. \u00a0I think I'm slightly positive on it, but I prefer the idea of directly raising kids.", "timestamp": 1362342984}, {"author": "Chris", "source_link": "https://plus.google.com/112938759017605010116", "anchor": "gp-1362343060246", "service": "gp", "text": "@Madeleine\n\u00a0If I make my genome public and someone uses it to create children of mine without my permission, am I required to pay child support as their father? \u00a0:)", "timestamp": 1362343060}, {"author": "Mad", "source_link": "https://plus.google.com/100533872344198336746", "anchor": "gp-1362343658607", "service": "gp", "text": "Right, DNA synth cost is exorbitant. On the other hand, if you share cell lines growing those is very easy. Such a defense hasn't been used in court as far as I know (and probably not for quite a while) -- not many people have publicly shared like this yet.\n
\n
\n[Sadly, almost all people with shared cell lines right now are unaware of it, thanks to how research has been conducted under traditional assumptions of privacy and de-identifiability of samples. And continues to be conducted, frustratingly. On the other hand, a lot of them are probably dead at this point... what rights families should have is another thorny issue!]\n
\n
\nI suspect IVF children is a more tangible risk for famous participants, e.g. George Church or Steven Pinker. But still very science fiction at this point.", "timestamp": 1362343658}, {"author": "David Chudzicki", "source_link": "https://plus.google.com/106120852580068301475", "anchor": "gp-1362371496893", "service": "gp", "text": "\"If/when insurance companies start to discriminate, they would almost certainly be doing so for much more well-established risks...\"\n
\n
\nAre you sure? Naively (knowing a bit about machine learning, a bit about auto/home insurance, and almost nothing about genetics) it wouldn't surprise me if very quickly life insurance companies could gather enough data to make predictions that lead to price differences for reasons no one can understand. (\"People with this combination of characteristics are 10% more likely to die each year, but we don't know anything about why.\")", "timestamp": 1362371496}, {"author": "David Chudzicki", "source_link": "https://plus.google.com/106120852580068301475", "anchor": "gp-1362371755108", "service": "gp", "text": "Jeff-- other people's genetic privacy might be a concern for you. Maybe you don't care about your own privacy, but care about letting the people genetically related to you maintain theirs. \n
\n
\nThis is really a concern with publicy of all types.", "timestamp": 1362371755}, {"author": "Jeff Kaufman", "source_link": "https://plus.google.com/103013777355236494008", "anchor": "gp-1362372551197", "service": "gp", "text": "@David Chudzicki\n\u00a0\"other people's genetic privacy might be a concern for you\"\n
\n
\nI wonder how many degrees of genetic separation it's good to ask permission for. \u00a0Immediate relatives only?", "timestamp": 1362372551}, {"author": "David Chudzicki", "source_link": "https://plus.google.com/106120852580068301475", "anchor": "gp-1362374299301", "service": "gp", "text": "We'd want to address the scientific (what do you risk revealing with what probability at what genetic distance?) and the ethical (for each answer above, how much do you care?) separately.\n
\n
\nI don't know anything about the science, but I suspect the answer is \"more than you'd think\". For example, I'm just now looking at the article Madeleine linked above and while maybe not directly related to that particular question, it's pretty amazing. ", "timestamp": 1362374299}, {"author": "David Chudzicki", "source_link": "https://plus.google.com/106120852580068301475", "anchor": "gp-1362374571902", "service": "gp", "text": "(\"This study shows that data release, even of a few markers, from one person can spread through deep genealogical ties and lead to the identifica- tion of another person who might have no ac- quaintance with the person who released his genetic data. The propagation of information through shared male lines amplifies the range of identification, allowing ~135,000 records to potentially target several million U.S. males.\")", "timestamp": 1362374571}, {"author": "Mad", "source_link": "https://plus.google.com/100533872344198336746", "anchor": "gp-1362403301950", "service": "gp", "text": "@David Chudzicki\n: \"predictions that lead to price differences for reasons no one can understand\"\n
\n
\nThese sorts of associations are often due to racial/ethnic ancestry (which are associated with socioeconomic associations, leading to health associations). Geneticists (somewhat euphemistically?) refer to this issue as \"population stratification\". Since racial discrimination is illegal and a little legal concern can go a long way, it seems likely insurance companies would start with variants with well-established health consequences.\n
\n
\nI don't have any strong opinions on how far one should go in discussing with relatives. I guess one might note that uncertainty can give a lot of privacy. For example, even if two parents reveal their genomes there's still uncertainty regarding which parts a child inherited (although not for mitochondria or Y-chromosomes).\n
\n
\nThe Personal Genome Project's ethical approach for this issue and others is to place strong emphasis on personal autonomy (it's your decision to piss off relatives). Related to that, participants must autonomously consent (i.e. no children, dementia, intellectual disability, or dead people).", "timestamp": 1362403301}, {"author": "Jeff Kaufman", "source_link": "https://plus.google.com/103013777355236494008", "anchor": "gp-1362403927182", "service": "gp", "text": "There's also the combination of the life insurance question with the genetic relations questions: could it make it more expensive for your relatives to get life insurance?", "timestamp": 1362403927}, {"author": "David Chudzicki", "source_link": "https://plus.google.com/106120852580068301475", "anchor": "gp-1362420136069", "service": "gp", "text": "@Madeleine\n\u00a0Oh, yeah. Oops! I should have thought of the race/ethnicity associations.\n
\n
\nBy the way, do you happen to know any of the authors of the deanonymization paper you mentioned?", "timestamp": 1362420136}, {"author": "Mad", "source_link": "https://plus.google.com/100533872344198336746", "anchor": "gp-1362422799288", "service": "gp", "text": "@David Chudzicki\n\u00a0Not personally, but I know people who know people. (e.g. I knew about the study before it was published.) Feel free to send me a message if you'd like assistance getting in touch with someone. Yaniv Erlich seems to be active on twitter (@erlichya).", "timestamp": 1362422799}]}