Half a croissant, on a plate, with a sign in front of it saying '50c'
h a l f b a k e r y
May contain nuts.

idea: add, search, annotate, link, view, overview, recent, by name, random

meta: news, help, about, links, report a problem

account: browse anonymously, or get an account and write.



Confidential public-access genome database

Find out what your chances of inherited diseases are without violating confidentiality
(+3, -3)
  [vote for,

I've looked for this, but it doesn't seem to be either halfbaked or baked. If it is, sorry.

The doctor-patient relationship is confidential. However, if someone is found to have an inherited disorder, this information is useful to their relatives, but they may not wish to disclose this information. Moreover, there are estranged and adopted children, people who think their fathers are not who they really are and so on. The wide availability medical statistics of the general population are not generally considered to be a breach of confidentiality. Other problems include insurance.
In the UK, each individual has a unique NHS number, and i presume that elsewhere others have something similar. There are also various genealogy projects which attempt to establish genetic relationships between people.

My idea, and i will frankly be surprised if this hasn't been proposed before, is this: All identifiable human genetic information which is linked to an individual via a unique identification number, as opposed to forensic information, should be registered on a database associated with that number. Any member of the public can then register an account with that database using their own identification number, in the case of this country the NHS number, evidence for identity and a means of establishing their relationship along with factors such as gender and lifestyle choices. The database then provides them with a risk assessment based on confidential information about relatives which is nonetheless anonymous because of the aggregation of the data.
I can see several problems with this. Very rare conditions would easily identify people, for instance recessive Robinow; a high probability of a particular condition would also sometimes do that. Another is obviously security, but then that's always a problem.

It would also be possible for two people to get the relevant probabilities for possible children by "mating" their data online with mutual consent.

nineteenthly, Mar 13 2009

Provoked by this http://www.bbc.co.u...1/2009_10_fri.shtml
(but not a new thought) [nineteenthly, Mar 13 2009]

Genetic Data troubles. http://www.pubmedce...d&pubmedid=15214961
zen_tom's anno reminded me of this article about a problem with storing genetic data, which comes completely out of the left field. Really it's about incompetent use of software rather than genetics but hey! [DrBob, Mar 13 2009]

UK government backs open source. http://news.bbc.co....hnology/7910110.stm
Yeah, right. [DrBob, Mar 16 2009]

Gattaca http://www.imdb.com/title/tt0119177/
Science fiction film. A decent stab at portraying a society where genetic science has gone mad. [DrBob, Mar 18 2009]

Your idea is bigger.... http://www.personalgenomes.org/
...but 100k is not bad for a start [loonquawl, Mar 18 2009]


       //Another is obviously security, but then that's always a problem//

Oh well, that's alright then! (-)
DrBob, Mar 13 2009

       I wonder what the database structure is for these genome databases - I mean, what datatype would you even use to store a genome or a chomosome? varchar(3,200,000,000) just isn't that practical, maybe they just put it all in a blob, or turn it into xml first, that they store in-situ to be unravelled by an xml parser on retrieval.
zen_tom, Mar 13 2009

       This is baked for diseases caused by genetic defects among orthodox Jews. It's operated by a large charitable organisation I can't remember the name of. You'd have your blood sampled as a child. Later, if you want to get married you'd submit your anonymous code number and that of your prospective partner to this organisation who would give you back a response which would just say whether there was a high risk of your children contracting diseases arising from genetic defects. As far as I know it's been very succesful.
hippo, Mar 13 2009

       Wow, that link's scary [DrBob] - I've seen accounts credited when they shouldn't have been on account of Excel's helpful data-guesswork features, I didn't imagine that it might "cause genes, including medically important ones, to be lost from view and which [have] contaminated even carefully curated public databases" - not a ringing endorsement...   

       Hi! It looks like you're analysing information for the human genome project! Do you want me to:
a) randomly destroy key pieces of data without your express knowledge?
b) propagate untold errors into the genome sequence itself, rendering worthless one of the great undertakings of man and potentially causing hideous and irreversible mutations and ultimately destroying all life on Earth?
c) write a congratulatory letter to Bill Gates?
zen_tom, Mar 13 2009

       Scary indeed! I spend a lot of work time helping people in my office to unravel the secrets of why their Excel spreadsheets aren't doing what they thought it should do, which is why I stumbled across that article.

I was accused of being an incredible nerd when I sent the link to my boss but it does bring out the science fiction writer in me! I wonder how much else of the world that we live in has been subtly perverted by unnoticed auto-complete errors in software packages? How many wars have been started? How many fantastic inventions have been lost? How many people disappeared from the records?
DrBob, Mar 13 2009

       Why a database? Couldn't you carry your genomic info on a personal USB key, encrypted and password protected? If you're looking to procreate with another keyholder, just walk up to a compatibility booth together, stick your keys in and wait for a detached, mechanical verdict. That way, no data is kept by any shadowy entity, the worst that could happen is you lose your key or forget your password, in which case you have to shell out the cash for another sequencing.
placid_turmoil, Mar 13 2009

       Seriously, [zen_tom], LOL! (Sorry for the smiloid).   

       [DrBob], i do realise i'm just saying that's a problem and i don't know what to do about it, but is it a greater problem than any kind of identity theft, nicking money or online crime of another kind? I didn't go into detail because i know little of online security other than what most people know, so whatever i said would be naive and inadequate. Concerning the Excel business, i'm rather concerned that the public sector uses proprietary software to that extent anyway as i don't remember voting for any of my VAT to go to Microsoft at any point (i don't pay income tax of course). [zen_tom], i didn't mean the entire genome, just a record of alleles associated with problems along the lines of alpha-one antitrypsin deficiency or a gene for dibasic amino acid mediator proteins in distal kidney tubules. Having said that, i don't think it'd be that hard to compress an entire genome. Many variations would be incompatible with life, so those can be ignored, many others wouldn't occur in humans and there are in any case more codons than amino acids, so those can be ignored too. I don't think it's an issue.
nineteenthly, Mar 13 2009

       [Placid_turmoil], i love the idea of memory sticks making sweet lurve (or is it just raw sex?)! I imagine the female version would have a socket and the male a plug. You should post that separately. It isn't quite the same as my idea though, since mine is not just about children but one's own risks.
nineteenthly, Mar 13 2009

       I don't get it: there seems to be a horse<>carriage question here. I mean if a family-member's family doctor discovers that that person has a predisposition towards <x> based on DNA then surely other family members' doctors can discover the same thing ?
FlyingToaster, Mar 13 2009

       Yes they can, but they may not have the same doctor or even be in the same healthcare system, and there may be no consent. There's a consent issue with the whole thing of course, but it's not quite the same as violating the right to withhold consent because the individuals involved are generally not identifiable. I'm not assuming everyone has their genome sequenced, just that certain alleles are corroborated in certain individuals. The day will doubtless come soon when it becomes cheap to do the whole thing, but that won't help when people have died and been cremated, for example.
nineteenthly, Mar 13 2009

       ok, that made my understanding of what you want to do even less. You don't think that Uncle Elroy upon learning he has a genetic predisposition towards Nintendoitis wouldn't tell his offspring/blood-relatives ? I think I'm missing a key bit of information here, sorry.
FlyingToaster, Mar 13 2009

       Whereas i don't always understand why, they don't always. There are a lot of people out there who don't know who their relatives are, for instance the descendants of sperm donors, children taken into care or adopted. Even if someone is willing to give consent, actually doing so is actively given and therefore takes time or can be accidentally omitted.
Here's an example of the kind of thing i'm talking about. An adopted child doesn't know who her parents are. She goes on to marry while pregnant. She then has a second child with her husband and gets pre-eclampsia. Pre-eclampsia is more likely to occur with first pregnancies by a father with blood group B, and has a genetic element. Only she knows the first child isn't his. Since she's adopted, she doesn't know much about her genetic background and therefore can't assess her risk of pre-eclampsia because only she knows the first baby wasn't his and she can't ask him what his blood group is. The database, on the other hand, knows what blood group his is, her genetic background and she can provide it confidentially with that piece of information. That would allow her to assess the risk more accurately. As it stands, all you seem to get in this situation is evasive medical staff and uninformed mothers who are at higher risk.
nineteenthly, Mar 13 2009

       //i don't remember voting for any of my VAT to go to Microsoft at any point// - I work for the Government and, I have to confess, had a cup of tea during work hours this morning. Next time I feel tempted to do this, I'll check with you first :-)
hippo, Mar 13 2009

       Yes, i understand the concepts of democracy, and i get where you're coming from. The point actually doesn't apply quite as well to this example as others. However, i don't understand why the government pays for products when there are free alternatives. Is it total cost of ownership?
nineteenthly, Mar 13 2009

       //if someone is found to have an inherited disorder, this information is useful to their relatives//   

       Is this supposed to read "...to have a genetic predisposition towards ..." instead of "...to have an inherited..."?
FlyingToaster, Mar 13 2009

       //Is this supposed to read...//Doesn't it amount to the same thing?
zen_tom, Mar 13 2009

       //amount to the same thing?//only if you don't see the difference between possibly getting breast cancer and actually having breast cancer. [19thly]s chain of events seems to be initiated by a person getting a disease as opposed to scientists discovering that a certain genome-sequence provides a predisposition for that disease.
FlyingToaster, Mar 13 2009

       // Is it total cost of ownership?// - yes, the economics of organisational use of open-source software aren't as simple as some people like to make out.
hippo, Mar 13 2009

       //only if you don't see the difference between// but I thought this idea was focussed on the likelihood of passing these traits on to your children - it doesn't matter whether you get the disease or not - if you are genetically predisposed - by inheritance - then, by inheritance, so might your children - by comparing your alleles with those of your intended partner, aren't you determining the overall probable likelihood of your children inheriting those same genetic predispositions from your pooled genetic makeup?   

       Or to put it another way, is it possible to become genetically predisposed to something through any mechanism other than inheritance? Sure, you may live in a pool of toxic ooze, but for the rest of us, it seems reasonable to suppose that your genetics are determined by normal genetic processes.
zen_tom, Mar 13 2009

       which is why I was asking for clarification... except for your mutant comment... pre-disposition towards comicbook superpowers ?
FlyingToaster, Mar 13 2009

       It isn't necessary to sequence an entire genome or have the clinical signs of a disease to be aware of a possible genetic problem. Alpha-one antitrypsin deficiency, HLA-specific antigens and blood groups are examples. Rhesus factor and to a lesser extent pre-eclampsia are pretty straightforward examples. However, clearly an individual with a completely sequenced genome would be quite valuable to her (or more likely his because of X-linking) relatives or potential relatives, and the existence of particular clinical pictures often provide a pretty good bet that someone has a particular genetic profile, for instance HLA B27. So, people having died of something would provide a pretty good bet in some cases and an assessment of probability would be feasible. So i'm not just referring to a chain of events initiated by someone having a disease, but it would include that.   

       It isn't just about children. Using the antitrypsin example, it might motivate someone to give up smoking or alcohol, and awareness of the adenomatous polyposis gene could lead to regular endoscopy to detect early signs of colon cancer.   

       And yes, it is possible to become genetically predisposed to something that way, for instance i'm aware of an example where it seems to have happened with peripheral neurofibromatosis. I don't think any radioactive spiders were involved though. In that sort of situation, this idea would be completely useless of course.
nineteenthly, Mar 13 2009

       still missing something here :/   

       Can't a person simply walk into a clinic, have their DNA sequenced and have that run against a medical database which includes genome-sequence issue flags ? There's a matter of professional standards here; it is to be assumed that the clinic making the request is doing so on behalf of a first-person inquiry (ie: not of the "I found this hair on my husband's jacket..." genre).
FlyingToaster, Mar 13 2009

       That could happen at some point, but i don't believe it's yet possible. I thought the cost of genome sequencing was still in the thousands. If it is, it's actually an argument for socialised healthcare because a rich person could benefit from the sequencing of a poor person's genome. Even so, i still think it's rather beyond the abilities of most medical practices. Am i wrong?
OK, i've got a year old reference to a hundred K dollars then and speculation as to a ton in four years time. That would probably make the idea redundant, but the mating memory stick thing would still make sense (though that's [placid_turmoil]'s idea, not mine).
Oh yes, another point about this. Having the genome doesn't tell you about its interaction with the environment. Epidemiology is still important. There could be, for instance, people who would metabolise a particular compound in a different way or a different rate but with which they've never come into contact. I don't think it's impossible that, say, aldehyde dehydrogenase has mutated in some long-standing Muslim populations and they wouldn't know about that unless they had their genomes done, but suppose there's a new drug out there with which nobody has yet come into contact, and people have different enzymes for it?
nineteenthly, Mar 13 2009

       then I am missing something... are you saying that when you hear on the news "blah blah blah DNA evidence" that each of those tests cost hundreds of thousands of pounds/dollars/whatever ?
FlyingToaster, Mar 13 2009

       No, i'm talking about doing a whole genome. Forensic DNA testing is different.
nineteenthly, Mar 13 2009

       // is it a greater problem than any kind of identity theft, nicking money or online crime of another kind?//

No, but to my mind the best way of avoiding identity theft is not to have an identity. Every time I see one of these bigger, better database ideas I imagine it in much the same way as I suspect that the crooks see it, as a big cake with ‘Eat me’ written on it.

//i'm rather concerned that the public sector uses proprietary software to that extent //

Me too, but most local government in-house IT expertise (by which I mean programmers as opposed to desk top support) has gradually been outsourced over the last twenty years. Lack of in-house expertise means that proprietary software suppliers who offer technical support for their applications are in the driving seat. Government is slowly waking up to the fact that software licensing for basic applications is a big, fat rip-off and has started making noises about Open Source solutions (linky) but, frankly, I don’t believe that it has the slightest intention of going down that path in any meaningful way.
DrBob, Mar 16 2009

       I have a lot of sympathy with that view (on ID). For a long time, i had no accepted means of providing evidence for my identity, but i eventually caved. I still do my tax return and books on paper alone. The question would seem to be, what can be done with that information which you wouldn't want? I can see a problem with insurance, extortion or hate crimes. Again, the issue is security and i don't know about it, so i don't know how to deal with that. Would it work if there was a whole network separate from the internet linking terminals only accessible in certain centres? For instance, would it be a question of getting access on demand but only in a health centre or hospital? That would allow genetic counselling to be given too.
On the open source and proprietary software issue, i'm well aware of that since several friends have lost jobs as a result, mainly COBOL programmers. I think it raises the issue of whether a generic approach to work is appropriate, but aside from that, i wonder if the place to break in is actually at compulsory education. In my position, i know nothing of English state schools, but i wonder if they just used open source, it would reduce total cost of ownership because everyone could be assumed to be familiar with particular open source applications. And the estimate of six hundred million is presumably just Whitehall?
nineteenthly, Mar 16 2009

       //Would it work if there was a whole network separate from the internet linking terminals only accessible in certain centres? //

Nope, because even if you isolate the system from the interweb that wouldn't prevent anyone from stealing the hardware nor prevent somebody from leaving their memory stick on the train.

I've no idea where they get the £600million figure from. They probably just made it up, but I can tell you that we spent about £750k on software licenses last year and we are a medium-sized local authority.
DrBob, Mar 17 2009

       OK, i'm now thinking a direct link between evidence for that exact genome, but i worry that this could still be cracked by means of lopping bits off people and using them for identity theft, or maybe just nicking a cup of coffee and wiping off the saliva.
nineteenthly, Mar 17 2009

       No, TERRIBLY dangerous idea!   

       Did you ever read those stories about google leaking it's search information? It was "anonymous," but just by cross referencing the search terms for almost any user, you could pinpoint ho it was from amongst everybody in he entire United States.   

       That's an anonymous 300 MILLION person pool, and the data was enough for it to not be at all anonymous. Now let's consider your idea, which involves an "anonymous" pool of about 20 people. It would be trivially easy for anybody using this to figure out exactly what the genetics of everybody in their family is. Just, no...   

       Also, insurance is automatically a problem as soon as individuals are able to learn about their own DNA codes. It doesn't have to get "leaked" to the insurance companies. Your rates will go up anyway, because anyone who finds out they are genetically healthy, will STOP BUYING INSURANCE. Thus, the insurance companies suddenly find themselves with a customer base that is oddly unhealthy. "How curious! Well, I guess we're going to have to jack up rates, then, huh?"
Smurfsahoy, Mar 17 2009

       I don't remember Google leaking anything, that was AOL wasn't it?
Spacecoyote, Mar 17 2009

       Now that I've got you feeling as paranoid as I do, nineteenthly, I feel that my work here is done! (llinky)
DrBob, Mar 18 2009

       Yes, it made me think of Gattaca too. Regarding insurance, isn't this an argument for public funding of healthcare? To someone in the UK, that probably looks obvious. For people in various other places, maybe not.
My attitude to insurance is generally that it's something you pay money into, never expecting to see that money again whatever befalls you futurally. My daughter recently had her bike nicked and yesterday someone suggested i claim it on insurance, and that had simply never occurred to me. I've always thought of insurance as a shredder into which you insert large denomination notes, so you do it when the law says you have to, not because you ever plan to claim on it.
Governments have large amounts of info on all of us, presumably so they can put it on memory sticks and leave them lying about in conspicuous public places. It's nice of them to be so committed to freedom of information, isn't it? Funny how they seem to do it more often with private stuff than the likes of defence plans (though yes, i know that does happen).
nineteenthly, Mar 18 2009

       //Governments have large amounts of info on all of us, presumably so they can put it on memory sticks and leave them lying about in conspicuous public places//
... it isn't treason if the government does it, apparently. Though I wonder if when I occasionally discuss my (usually criticism of my) government's policies with people of other constituencies if I am committing that offense.

       umm do you have a link explaining the difference between forensic DNA and genome DNA (Wikipedia didn't help)
FlyingToaster, Mar 18 2009

       What i mean is, it's possible to find similar lengths of DNA without having to know what the DNA actually does, if anything, so if all you're looking for is to find out if someone has been at a crime scene you just need to compare their DNA to the DNA found there, but if you're trying to find actual genes and find out what they do, it's a much bigger job.
nineteenthly, Mar 18 2009

       umm okay, at some point I'll have to get a primer on this stuff (nvm)
FlyingToaster, Mar 18 2009


back: main index

business  computer  culture  fashion  food  halfbakery  home  other  product  public  science  sport  vehicle