Interesting. I do worry about the security implications, especially after watching this talk[0] about the security of Estonia's e-voting system. The government does not seem to have a wonderful sense of how to handle information securely, and does not take to criticism of its security protocols kindly. Here[1] is a tweet from the Estonian CIO in response to questioning about the video.
This is the same government who can't remember that RSA keys generally use positive numbers? https://crbug.com/532048
The idea in theory is interesting, but for something even more high-profile than national identity cards, please hire some folks who can triple-check your work....
I didn't see anywhere in the article mention the ability (or lack of) to use DNA for criminal prosecution. Does anybody have more information about that?
"Estonian Human Genes Research Act governs the activities of the biobanking project: It establishes anonymity in clinical research, enables donors to decide which studies they want to participate in, and gives donors full control over who has access to their data. By default, a donor’s doctor is the only other person who can look at his or her genetic information through the portal"
In Estonia, if you meant relative to this article, the DNA is anonymous and it is implied unusable in court. If you are referring to the recent article about invalidity of DNA evidence in court cases, IIRC it was procedural but I couldn't find much more on it, and I too would be interested.
THe article notes ~$5K to sequence a genome, it sounds like we are pretty close to ~$1K genome sequence[0] going into production. This is a great thing.
The privacy concerns surrounding this topic are very real and very complicated. It took me a long time to make up my own mind, however, I am convinved this is a great step forward. Having access to the sourcecode of 8 Billion github repos, if you had the technology to process the data, would make you quite good at fixing bugs so to speak.
The value of a genome will be quite high in the future, and if we allow the data sets to be public this will cut out many IP issues. Anonymizing the data is fundamental, as is an open process and total transparency around collection, storage and the resulting sequence. All this data needs to be public and free.
The value of rare genomic anomolies will be evident in the near future, and having public data is a way to combat unethical testing or information siloing.
> Can you elaborate on what do you think would be the great step forward?
As someone with little knowledge of the field or the science involved, my ideas are largely borrowed from others much smarter than myself. Steve Jurvetson, a VC from DFJ who regularly talks about this field, engineering solutions through evolution, and Gordon Moore. Craig Venter was a fundamental force in sequencing the entire human genome this research made companies like Anne Wojcicki's 23AndMe possible. They are the people who will likely provide the true merit of this endeavor.
My thoughts:
* Early diagnosis. This is one of the most obvious ones. With companies like theranos providing quick testing equiptment and the price of sequencing plummeting, having gentic evidence to compare for risk factors and real-time blood tests, the ability to diagnose things early would be amazing.
* Cures. Broadly, the ability to read, write and interpret genetic sequences would allow for personalized medicine and curing genetic illnesses one was predisposed to.
* Adaptation. If we were able to have access to a dataset so massive, we were able to uderstand and curate genes. Outside of basic gene therapy and medacine we could actively enhance humans. This could possibly provide electronic interface with our own wetware, the extension of life, or otherwise change aspects of the human condition that were not desirable.
* Creation. This is controversial, so I will leave it there.
There are so many ways even with current technology that would provide positive things if we had this data and the tools to analyze it. I am excited for this. Similar to AI which will bifurcate between extinction or utopia, this has quite a large upside but also some substantial negatives.
I am on a similar boat as you are with respect to the prospect of genomes revolutionizing health and the necessity and advantage of making it available openly (or at least to the researchers). Right now, most of the differences in humans tend to be more biological rather than ideological. The one thing that bothers me is - are our ideological differences just because of our biological differences? It just goes back to the question of free will. If there is no free will and we are just driven by what our genes dictate, all of us end up being very similar like machines. Then we can only be different by choosing different things intentionally. Are we really mature enough to do that? I am not sure if this belongs to the topic here but something I keep thinking once in a while and would like to know what others here think.
Self, intentionality and free will are mental concepts, not external objects to be observed from a third person perspective and studied like atoms and force fields. When trying to apply physics to psychology, such as trying to relate determinism and free will, a mixup of language domains occurs. Free will has a first person perspective (subjectivity) that places it on a different domain than physical forces and particles which can be observed externally.
The weird thing about DNA and genes is that the data, should you have access to it in it's original form, as samples of biological material, is never anonymous.
The techniques that effectively anonymize genetic information, typically render it as an unusable data set for practical purposes.
The only thing that makes DNA anonymous is constructing an unrelated set of hash keys as lookups, and then using these coded tokens as decoupled representational references to a remote system that mitigates requests. Even then, even without access to the direct data, the traits expressed COULD still accidentally identify certain members of the set very easily.
How many albinos are in the set? How many male albinos? How many male albinos with blood type O positive? Oh look, it's Joe and it turns out he's a carrier for this other embarrassing genotype. Whoops, secret's out and everyone knows, and there's no going back!
So phenotypes can still identify people inadvertently, and then provide opportunities for the enumeration of members, through process of elimination, inherently weakening what might become a theater of anonymity.
But more terrible still, is, let's just say they give the raw genetic data up. Say you have a juicy tidbit of information you pulled from the collection of samples, and you think you know who it is. So you surreptitiously snag a hair follicle from them. It's them all right, and you can prove it conclusively. Maybe your a nice guy and you won't misuse this sort of advantage, but is that true for everyone at large, throughout society?
With DNA, genomes and genetic evidence, you kind of just have to keep saying to yourself "this is not anonymizable, it will never be anonymous, we cannot truly de-identify anyone's DNA" and the sooner we come to terms with that, the better we'll get at dealing with this sort of data.
Re "anonymizing DNA": In a sense, isn't an individual's DNA sequence the closest thing to a perfect and intrinsic identifier? Only fails to work in the case of identical twins ...
DNA could be used for nefarious purposes. If eugenicists like the Nazis could have stolen Estonia's DNA records, they could have more efficiently committed genocide.
I can't find the story now, but there was one country which kept on file the religion of each citizen. Nazis gained control of the building with those records. Some rebels were successful in burning those files, saving many people's lives.
The historical record is full of key themes and tropes: our warlike behavior as whole, bias for tribalism, hatred and discrimination of minorities and society's propensity to rally behind those things.
To ignore the recent abuse of such data is ignoring the potential abuse DNA collection could enable.
22 comments
[ 2.4 ms ] story [ 59.7 ms ] thread[0] https://www.youtube.com/watch?v=JY_pHvhE4os [1] https://twitter.com/taavikotka/status/612281286011568128
The idea in theory is interesting, but for something even more high-profile than national identity cards, please hire some folks who can triple-check your work....
In Estonia, if you meant relative to this article, the DNA is anonymous and it is implied unusable in court. If you are referring to the recent article about invalidity of DNA evidence in court cases, IIRC it was procedural but I couldn't find much more on it, and I too would be interested.
The privacy concerns surrounding this topic are very real and very complicated. It took me a long time to make up my own mind, however, I am convinved this is a great step forward. Having access to the sourcecode of 8 Billion github repos, if you had the technology to process the data, would make you quite good at fixing bugs so to speak.
The value of a genome will be quite high in the future, and if we allow the data sets to be public this will cut out many IP issues. Anonymizing the data is fundamental, as is an open process and total transparency around collection, storage and the resulting sequence. All this data needs to be public and free.
The value of rare genomic anomolies will be evident in the near future, and having public data is a way to combat unethical testing or information siloing.
[0]http://www.nature.com/news/is-the-1-000-genome-for-real-1.14...
As someone with little knowledge of the field or the science involved, my ideas are largely borrowed from others much smarter than myself. Steve Jurvetson, a VC from DFJ who regularly talks about this field, engineering solutions through evolution, and Gordon Moore. Craig Venter was a fundamental force in sequencing the entire human genome this research made companies like Anne Wojcicki's 23AndMe possible. They are the people who will likely provide the true merit of this endeavor.
My thoughts:
* Early diagnosis. This is one of the most obvious ones. With companies like theranos providing quick testing equiptment and the price of sequencing plummeting, having gentic evidence to compare for risk factors and real-time blood tests, the ability to diagnose things early would be amazing.
* Cures. Broadly, the ability to read, write and interpret genetic sequences would allow for personalized medicine and curing genetic illnesses one was predisposed to.
* Adaptation. If we were able to have access to a dataset so massive, we were able to uderstand and curate genes. Outside of basic gene therapy and medacine we could actively enhance humans. This could possibly provide electronic interface with our own wetware, the extension of life, or otherwise change aspects of the human condition that were not desirable.
* Creation. This is controversial, so I will leave it there.
There are so many ways even with current technology that would provide positive things if we had this data and the tools to analyze it. I am excited for this. Similar to AI which will bifurcate between extinction or utopia, this has quite a large upside but also some substantial negatives.
The weird thing about DNA and genes is that the data, should you have access to it in it's original form, as samples of biological material, is never anonymous.
The techniques that effectively anonymize genetic information, typically render it as an unusable data set for practical purposes.
The only thing that makes DNA anonymous is constructing an unrelated set of hash keys as lookups, and then using these coded tokens as decoupled representational references to a remote system that mitigates requests. Even then, even without access to the direct data, the traits expressed COULD still accidentally identify certain members of the set very easily.
How many albinos are in the set? How many male albinos? How many male albinos with blood type O positive? Oh look, it's Joe and it turns out he's a carrier for this other embarrassing genotype. Whoops, secret's out and everyone knows, and there's no going back!
So phenotypes can still identify people inadvertently, and then provide opportunities for the enumeration of members, through process of elimination, inherently weakening what might become a theater of anonymity.
But more terrible still, is, let's just say they give the raw genetic data up. Say you have a juicy tidbit of information you pulled from the collection of samples, and you think you know who it is. So you surreptitiously snag a hair follicle from them. It's them all right, and you can prove it conclusively. Maybe your a nice guy and you won't misuse this sort of advantage, but is that true for everyone at large, throughout society?
With DNA, genomes and genetic evidence, you kind of just have to keep saying to yourself "this is not anonymizable, it will never be anonymous, we cannot truly de-identify anyone's DNA" and the sooner we come to terms with that, the better we'll get at dealing with this sort of data.
https://en.wikipedia.org/wiki/Estonian_alien%27s_passport
I can't find the story now, but there was one country which kept on file the religion of each citizen. Nazis gained control of the building with those records. Some rebels were successful in burning those files, saving many people's lives.
To ignore the recent abuse of such data is ignoring the potential abuse DNA collection could enable.
It's also been used to catch at least one murderer of a celebrity, but generally is not used for crime such as rape cases.