5 ms·

How to Identify Almost Anyone in a Consumer Gene Database

70 points by extraterra 8y ago

tinbad 8y ago
I recommend anyone interested in genome exploits to watch this talk at DEFCON by John Sotos[1] about why you should be very hesitant in giving away your DNA. [1] https://www.youtube.com/watch?v=HKQDSgBHPfY https://www.youtube.com/watch?v=HKQDSgBHPfY
- thedailymail 8y ago
  I'd heard about that talk before but never watched it – scary stuff! Many thanks for sharing the URL.
Real_S 8y ago
>New techniques ... may soon make the anonymity of their customers’ DNA impossible to safeguard I strongly disagree. >He suggested it might be wise to encrypt genetic data to protect personal information. Maybe we can encrypt DNA molecules themselves! Check out GeneInfoSec
- asaph 8y ago
  >> New techniques ... may soon make the anonymity of their customers’ DNA impossible to safeguard > I strongly disagree. Can you please share why?
okket 8y ago
See also previous discussion about this topic from 12 hours ago: https://news.ycombinator.com/item?id=18195827 https://news.ycombinator.com/item?id=18195827 (60 comments)
tgb 8y ago
Largely unrelated but MPEG of mp3 fame has just released MPEG-G, a compressed file format for genomic data. Predictably, it's on track to be saddled with patents. Imagine needing proprietary software to view your own genomic data - or for your new doctor to not have the software to read it! Here's one take worth reading on it: http://datageekdom.blogspot.com/2018/09/mpeg-g-ugly.html?m=1 http://datageekdom.blogspot.com/2018/09/mpeg-g-ugly.html?m=1 Also there's already a compressed data format called CRAM for this that has some traction and for which MPEG conveniently failed to compare itself to in their paper.
- jkbonfield 8y ago
  To be "fair", their latest revision of the paper includes some figures from a 4 year out of date version of CRAM, which is an improvement on the 10 year old outdated format they used initially. ;-) Disclaimer, that's my blog and I'm the primary author of the newer version of CRAM.

70 points by extraterra 8y ago

I recommend anyone interested in genome exploits to watch this talk at DEFCON by John Sotos[1] about why you should be very hesitant in giving away your DNA. [1] https://www.youtube.com/watch?v=HKQDSgBHPfY https://www.youtube.com/watch?v=HKQDSgBHPfY

thedailymail 8y ago
I'd heard about that talk before but never watched it – scary stuff! Many thanks for sharing the URL.

Real_S 8y ago

>New techniques ... may soon make the anonymity of their customers’ DNA impossible to safeguard I strongly disagree. >He suggested it might be wise to encrypt genetic data to protect personal information. Maybe we can encrypt DNA molecules themselves! Check out GeneInfoSec

asaph 8y ago
>> New techniques ... may soon make the anonymity of their customers’ DNA impossible to safeguard > I strongly disagree. Can you please share why?

okket 8y ago

See also previous discussion about this topic from 12 hours ago: https://news.ycombinator.com/item?id=18195827 https://news.ycombinator.com/item?id=18195827 (60 comments)

tgb 8y ago

Largely unrelated but MPEG of mp3 fame has just released MPEG-G, a compressed file format for genomic data. Predictably, it's on track to be saddled with patents. Imagine needing proprietary software to view your own genomic data - or for your new doctor to not have the software to read it! Here's one take worth reading on it: http://datageekdom.blogspot.com/2018/09/mpeg-g-ugly.html?m=1 http://datageekdom.blogspot.com/2018/09/mpeg-g-ugly.html?m=1 Also there's already a compressed data format called CRAM for this that has some traction and for which MPEG conveniently failed to compare itself to in their paper.

jkbonfield 8y ago
To be "fair", their latest revision of the paper includes some figures from a 4 year out of date version of CRAM, which is an improvement on the 10 year old outdated format they used initially. ;-) Disclaimer, that's my blog and I'm the primary author of the newer version of CRAM.