Page 1 of 2 12 LastLast
Results 1 to 10 of 13

Thread: What are the ACTUAL risks of DNA testing/sequencing?

  1. #1
    Registered Users
    Posts
    269
    Sex

    What are the ACTUAL risks of DNA testing/sequencing?

    In the wake of the "DNA Friend" spoof-site and recurrent aspersions cast over some testing companies, I thought we could try to have a facts-based discussion of actual risks. I would prefer not naming actual testing companies, because how easily that turns into unfounded bashing & defensive attitude. In turn I'll try to avoid calling things a conspiracy theory, but honestly casting accusations and suspicions without evidence is pretty low.

    I expect, if this discussion takes on, it'll be all over the place anyway... but I'll bring up what I consider to be one of my top concerns. (Also, yes, I know the wisdom of posting nearly book-length treatises is questionable, but I wanted to get most salient points into this).

    First is information security; many companies have already had their first security breaches, and if it occurred before GDPR odds are you never heard about it. I've personally discovered one genetic data leak by one clinical sequencing company, that was fixed afterwards, and the company shall remain nameless. Others have come up in news as well. This is particularly problematic, because DTC providers are likely to hold genetic data & customer information where they can be linked to each other. On the other hand, contrary to general belief, individual genome sequences without voluntarily given disease or relative information hold very little value and are easy to come by, for example the Haplotype Reference Consortium HRC alone holds nearly 40.000 human genome sequences available to all qualifying researchers (http://www.haplotype-reference-conso...pating-cohorts).

    General security theory says that information leaks can never be fully prevented, just made very hard, and this is doubly true if the attacker is a government actor, who are likely to have just industrial espionage agents inside competing countries operations. (https://www.sciencedirect.com/topics...rial-espionage) This bears particular relevance to demands to make customers DNA data more accessible online, since what isn't online can't be "hacked"; nevertheless, most of us likely prefer the added convenience of this, and are going to have the data online for further analysis anyway. A brief mention of physical security is also in order, since for example most courier companies (DHL, UPS, FedEx etc.) do not appear to identify recipient (Personal experience) allowing an even easier intercept. Consequently, the risk is always present, and with every company.



    While personal information is private and deserving of protection, what the exact effect of this risk is, is harder to quantify. Your identifying information and credit-card details are likely MUCH bigger target than the genetic data! In the case of industrial espionage for example, individual DTC customer is likely to be affected. Those who object to DTC testing frequently point out how inconsequential the data received supposedly is, and it's likely true that for the vast majority of the test-takers the results have nothing that isn't already reasonably public knowledge. In this context it's noteworthy that James Watson (of DNA structure co-discovery fame) published his whole genome, with the exception of his ApoE gene, associated with Alzheimer's disease. (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2986051/). But unless you're a very public figure, few people are going to be interested in your medical DNA results or unacknowledged children.

    In the USA (and other countries frequently have something similar), the Genetic Information Nondiscrimination Act prevents most harmful consequences of medical DNA information, and some states are moving even further by seeking to prohibit use of genetic information in life-insurance. If insurance companies were to use a specific DNA test they would almost certainly require their own DNA test to standardize on the results and remove the chance of submitting someone else's sample. As it stands however, they're already getting your genetic test results if you took them through medical channels instead of DTC in a process known as medical underwriting! (https://www.miamiherald.com/news/pol...227403299.html). They certainly would not use hacker-leaked data.

    Indeed, one of the largest risks is likely the reverse one: Increased availability of genetic testing services, and public knowledge of them makes it an attractive target for online scammers, like online porn-site scams have become. Except poorly written e-mails claiming to have your DNA data and threatening to reveal whatever secrets they're holding to hit your mailboxes in the future. No DNA tests necessary, because actually the threat or revealing the secrets will be all the more convincing if you haven't seen what secrets it holds. Inbred parents? Mental degradation? Unknown son from that affair they had while married? A few people are going to be convinced they have all of these on the basis of an e-mail. (https://www.theguardian.com/technolo...tortion-webcam)

    Of course, the effort needed to actually shift through people finding these (Which in most cases wouldn't be secret at all) & find current contact information to them means scammers are unlikely to do it for real and just see who falls for it. Of course, on a deeper ethical level one wonders if this violates that assumption of voluntarily consent built into all genetic testing services, if popularity of DNA testing leads to people needing to pre-emptively test just to know what's NOT in their genes?

    Next up.... technical & judgemental errors of genetic testing. Next year, maybe...
    Last edited by Donwulff; 04-25-2019 at 02:40 PM.

  2. The Following 6 Users Say Thank You to Donwulff For This Useful Post:

     CannabisErectusHibernius (05-26-2019),  DMXX (05-09-2019),  euromutt (04-25-2019),  karwiso (05-01-2019),  sktibo (04-25-2019),  timberwolf (05-13-2019)

  3. #2
    Registered Users
    Posts
    269
    Sex

    One reasonably quick fix to a minor issue mentioned in that is that companies which offer offline delivery of data should provide them encrypted, with the encryption key provided on their authenticated web-service to ensure the data won't end in wrong hands at customs, receiving or even just at the home it's sent to. I'd hate to say "optional" because usually that would mean it gets skipped, especially in case someone impersonates the tested to request offline delivery, but I admit encryption would add extra technical challenge for some. For the extra paranoid, this could be expanded to ALL results being delivered off-line, and more companies could consider offering the option for those who are held back by worrying about privacy, although in the 21st century I suspect it's impossible to avoid the data ever being loaded on an Internet connected machine.

    Of course, with hand-held sequencers already in existence, this level of security seems to be quickly becoming useless. As Google Ventures chief (After me *cough*) said in 2015, your genome isn't really secret: https://www.beckershospitalreview.co...-think-so.html "Mr. Maris said anyone who wanted to access anybody else's genome would simply have to collect any piece of genomic material, be it a hair or even a cup with traces of someone's saliva. Hacking a server to access genomic data is not necessary, he said." (Along with the point that your other personal data is much more valuable & harmful). Of course, it doesn't hurt to give people a little peace of mind, even when in the long run it will be meaningless. Instead, we should be concerned about building a society where the thinning privacy of genetic data isn't an issue; laws like the GINA are a step in the right direction, although ultimately they can't affect people's internal prejudices.
    Last edited by Donwulff; 04-25-2019 at 07:13 PM.

  4. The Following User Says Thank You to Donwulff For This Useful Post:

     DMXX (05-09-2019)

  5. #3
    Registered Users
    Posts
    30
    Sex
    Location
    Vermont, USA
    Ethnicity
    NW European
    Nationality
    American
    Y-DNA
    R1b-Z295*
    mtDNA
    U2e1j

    I believe that the Harvard Personal Genome Project did a very thorough job listing semi-plausible-at-the-time potential risks of sequencing. It broke them down into three general categories of risks associated with public disclosure of your sequence (from learning undesirable genealogical facts, to revealing information about your relatives, to someone fabricating DNA to match yours and placing it at a crime scene), risks associated with the creation of cell lines from your donated tissues (from you may be cloned to your stem cells may be made into a human-animal chimera), and risks associated to your receipt of sequencing data (from poor data quality to medical errors due to care sought on the basis of your sequencing results).

    See https://my.pgp-hms.org/static/PGP_Co...ne_stamped.pdf , starting at page 13.
    Last edited by bjp; 04-25-2019 at 06:59 PM.

  6. The Following User Says Thank You to bjp For This Useful Post:

     DMXX (05-09-2019)

  7. #4
    Registered Users
    Posts
    269
    Sex

    Tbh, I think many of those are NOT realistic risks, and of note specifically "6.2 Risks Associated with the Creation of Cell Lines From Your Tissues" actually concerns cell lines from biological sample, not the sequencing itself. It's hard to imagine where any of this could be even theoretical issue with saliva (Or to the extent it was, the same issue would exist any time you spit or leave a chewing gum ball or whatever). To be honest, there was 2013 spectacle of an artist creating facial reconstructions of DNA left in public: https://edition.cnn.com/2013/09/04/t...res/index.html Although I believe this was more a publicity stunt than a real thing, especially considering facial reconstructions from DNA are still very, very inefficient: https://www.biorxiv.org/content/10.1101/185330v3

    Many if not all of the DNA sequence specific risks of the PGP paper can actually be found on most DTC genetic testing companies consent/terms of service pages as well. Of course, they do not exactly stress the point that they could be subject to information leak/hacking, or accidentally send the data to wrong person.

  8. The Following 2 Users Say Thank You to Donwulff For This Useful Post:

     bjp (04-25-2019),  DMXX (05-09-2019)

  9. #5
    Registered Users
    Posts
    269
    Sex

    Also, I want to stress that the whole point of PGP is that the genome sequences *will* be made public. So that shifts the risk model from "What are the most likely ways of unintended disclosure of your genome" into "Now that your genome is widely publicized, what are the potential risks?" That's not to say that they aren't both potentially applicable, just different points of view - also, in the case of PGP, since that's the actual "consent document", covering their asses in case of the most unlikely of consequences. As evidenced by GINA (2008), the Maris quote, and me preaching the same even before then, in 2015 when the PGP consent document was written it was already visible that the widespread availability and advancement of sequencing technology would erode genetic privacy in such a way that people could not keep their DNA secret in future. Regardless, it's a good and necessary thing they enumerate the risks to participants.

    Some may even remember the infamous case, already in 2002 way before all this DTC genomic testing hulabaloo, of millionaire Steve Bing and his discarded dental floss - used by a private investigator to prove his paternity to demand child support. https://www.theguardian.com/educatio...ighereducation (Him being a millionaire, this led to the first law against surreptitious DNA gathering, albeit of course a law is unlikely to stop people with criminal intent) Certainly, there ARE risks involved with unwanted disclosure of genetic information, particularly for the rich and famous, but they're not uniquely - or likely even predominantly - created by DTC genetic testing of ourselves. This kind of surreptitious or "crime-scene" DNA sampling seems fairly common tactic for private investigators and law enforcement nowadays: http://www.abajournal.com/magazine/a..._if_you_stole_ Naturally this doesn't *reduce* the risk from testing ourselves, but it puts it in context and illustrates how our DNA has not really been a "secret" known only to us if we choose so for a long time now. To say nothing of the DNA we share with out relatives.

    Of course, the "learning unwanted information" risk always stands, albeit we can only hope for voluntary, fully informed consent (Which, as noted in my opening post, potential blackmailing scams threaten). Most people on online discussion forums don't seem to consider potential unwanted information an issue, but they're quite real: https://www.theatlantic.com/science/...ernity/562928/ It would certainly be folded in with "errors of technology & judgement", ie. that no kind of genetic testing the unerring, unwavering oracle that many seem to perceive it to be. Often they're more like the capricious oracles of Greek legends making prophecies that are mis-interpreted, otherwise useless or self-fulfilling, although there are obviously gems among them, as well and the ability to interpret genetic data is constantly improving. https://geneticliteracyproject.org/2...na-sequencing/ (Which is one opinion piece, though).
    Last edited by Donwulff; 04-25-2019 at 10:29 PM.

  10. The Following User Says Thank You to Donwulff For This Useful Post:

     DMXX (05-09-2019)

  11. #6
    Registered Users
    Posts
    3,820
    Sex
    Y-DNA
    R1b
    mtDNA
    H

    Don, you are minimizing the risks of making your whole genome data available as well as sending DNA samples into areas you don't know, but your point of view is legitimate.

    I was not aware of this DNA Friends web site. https://dnafriend.com/about

    Are these people in the "About" real or are they spoofs too? What a wierd thing to provide a parody of on a web site. If these people are not real and someones' name is being used in this spoof that's pretty bad.

    I can find Dr. Kenner's in Linkedin and FB.

    Don, are you a biologist? or something? I'm not. I have to go on what I read but if you have a particular specialty here that would be good to know.

  12. #7
    Registered Users
    Posts
    269
    Sex

    I'm data scientist with bioinformatics background. Regardless, these are points upon which experts disagree, which is fine, I'm specifically asking for viewpoints and they should be weighted on reason & references, not on appeal to authority of pseudonymous online posters.

    I'm not sure where you get "minimizing the risks", although it is my contention that the risks that people seem to most often cite (Which DNAFriend is a good example of) are way overblown, while some other risks like unintended disclosure/hacking that have panned out several times already aren't even talked about. Even so, it would be a worthwhile discussion what IS the actual harm of having your genome public; because let's face it, if anyone *wants* your genetic information they can already get it. And the law is generally behind, patently criminal for hospital or DTC company to sell your DNA data to China without your consent, but picking someone's discarded paper cup and having the DNA sequenced for paternity/disease information is in many if not most cases perfectly legal.

    And yes, DNAFriend is a satire, it's concerning to me that most people even on Anthrogenica don't seem to be able to tell it's fake right away. They're an Elon Musk bankrolled parody company, although he's distanced himself from them specifically because he doesn't want to be associated with their view. In an interview they gave to The Atlantic, they made it pretty clear the site is intended as a serious guerilla media "attack" on DTC genomics. Yet even so (Or, perhaps specifically so) it's just re-cycling old tired arguments that can be found on any online forums where the topic is discussed without even as much as providing a link for reference or further information.

    Only somewhat related to topic at hand, but on the topic of genomic medicine, how integral but lawsuit-prone sequencing has become to healthcare, and how beneficial having your DNA sequence around is: https://www.sciencemag.org/news/2019...ent-far-behind (Although I wish they'd put in even a word about the opposite case of genetic mis-diagnoses due to false positives, now it comes across as overly pro-sequencing piece).

  13. The Following 2 Users Say Thank You to Donwulff For This Useful Post:

     DMXX (05-09-2019),  Judith (05-25-2019)

  14. #8
    Registered Users
    Posts
    84
    Location
    ancestry Irish, Scot, Eng (Devon,Somerset,Lancashire,Cheshire, Shrop,Essex)+ Crossen(Oder)-Gruenberg
    Ethnicity
    Celt? Vik? East-Ger Sorb?
    Nationality
    Y hu R1b/U106/S1688/Y9825
    Y-DNA
    f.R1a/L1029/YP263
    mtDNA
    K1a26,HV0-T195C

    Quote Originally Posted by Donwulff View Post

    risks like unintended disclosure/hacking that have panned out several times already aren't even talked about.

    Even so, it would be a worthwhile discussion what IS the actual harm of having your genome public; because let's face it, if anyone *wants* your genetic information they can already get it.

    Noone can possibly know the actual risk, no one knows what the future will bring, but possible risks, in my view, apart from hacking by organised crime or bureaucracy are mainly from those future risks with the unforeseeable but arguably, based on history, expected, madmen at the helm situations.

    Hacking- 1 . merely to obtain some DNA sequences.. no, I can't think of any reason for this, but perhaps I lack foresight.
    2. In paternity cases or criminal cases... maybe a reason there. Can anyone see criminal gangs (like say Mafia) using DNA to locate people, the police have already started. (although you can't as yet say there are any cases of known hacking. I'm sure they will occur if deemed necessary sometime, somewhere and by somebody in the future). The question is when it occurs, does it matter to some relative in the future?

    Future Hitler/Stalin types who for some warped reason want to commit some kind of mass genocide or similar may use it but they could easily obtain it anyway? I can't imagine any scenario where the govt would want to eliminate all descendants of joe-blow who just ran his WGS or even his FTDNA chip.. but maybe my imagination is too limited. I think Hitler would have used it , if he had had it..so that was my main concern.

    To me, I was concerned, But try as hard as I can, I always can't beat the argument of " let's face it, if anyone *wants* your genetic information they can already get it. "

    So it's just an argument of aking it as safe as possible for future generations. Which boils down to..why would any say Hitler type or Mafia type etc in the future, say in 150 years time, want to compare my DNA with people of that time and maybe provide some selective treatment based on their relationship with me(or more likely based on their relationship with a genetically related person, or a grouping of which I am part of ..like a group of SNP's which someone deems to be necessary to wipe off the face of the earth) ..and there my imagination draws a blank.

    Thus, I am now finally running a couple of WGS.
    As far as getting unwanted info is concerned, I got that from one of my first DNA tests on FTDNA, when someone came up as a relative, from part of the family whom the family had always said was only related by marriage. Someone you wouldn't want to be genetically related to, even remotely. (but I guess a lot of people have those somewhere... you say you haven't? sigh). It's a good reason not to look at your matches and merely your origins Then again origins are giving surprises too.. I think I like tracing the ancient DNA movements better
    Last edited by Jan_Noack; 05-10-2019 at 08:26 AM.

  15. #9
    Gold Class Member
    Posts
    109
    Sex
    Location
    Cambridge, MA USA
    Ethnicity
    Polish/British Isles
    Nationality
    U.S.
    Y-DNA
    R-A9185
    mtDNA
    H1

    Poland England Ireland Munster
    My DNA test results including raw data from 23andMe, Ancestry, Dante Labs (WGS) and FGC (Chromium Long Read) are on Google Drive and available to the public at this link:
    https://drive.google.com/open?id=1_x...hcNbyizx1a39F3
    As the Readme file in that folder states:
    You are welcome to download the files in this directory and share them with others.
    Sincerely,
    Patrick O’Keefe
    I created that folder and made it public September 2018.
    Obviously privacy issues, to the extent I am aware of them at this time, do not worry me much.
    Currently I am happy to serve as a test case, if anything awful happens to me as a result of making my DNA test results public, we will have that much evidence that it is a bad idea
    YFull: YF14620 (Dante Labs 2018)

  16. The Following User Says Thank You to pmokeefe For This Useful Post:

     Jan_Noack (05-11-2019)

  17. #10
    Registered Users
    Posts
    30
    Sex
    Location
    Vermont, USA
    Ethnicity
    NW European
    Nationality
    American
    Y-DNA
    R1b-Z295*
    mtDNA
    U2e1j

    Quote Originally Posted by pmokeefe View Post
    My DNA test results including raw data from 23andMe, Ancestry, Dante Labs (WGS) and FGC (Chromium Long Read) are on Google Drive and available to the public at this link:
    https://drive.google.com/open?id=1_x...hcNbyizx1a39F3
    As the Readme file in that folder states:

    I created that folder and made it public September 2018.
    Obviously privacy issues, to the extent I am aware of them at this time, do not worry me much.
    Currently I am happy to serve as a test case, if anything awful happens to me as a result of making my DNA test results public, we will have that much evidence that it is a bad idea
    Cheers to a fellow risk-taker!

    I've done the same, but I'm using data storage at Open Humans to host as much as I can. My Veritas WGS is released to the public domain through the PGP and hosted there, along with ClinVar and GET-Evidence reports based upon it, also available to the public with no login needed. My FTDNA Sanger-sequenced mtDNA is KU530226 on GenBank (cheers to YFull for leapfrogging PhyloTree and codifying U2e1j in public sources).

    On Open Humans, I have my 23andMe v3 raw data, FTDNA Y-STR-67 data, American Gut Project five-site microbiome sequencing results, my Veritas WGS, all the 3rd-party analysis FGC ran on my Veritas data (lobSTR, yKnot, snpEff, VEP), and a link to my Genevieve report (an OpenHumans based VCF variant analysis tool) and GenomiX report too but that site appears down at the moment. My private chrY SNPs are openly viewable at the R1b Big Tree.

    This has all been out there for free public download by anyone with no restrictions for more than five years, I have not had a single adverse result. Anyone who wants any of it for any reason can get it all from there.

  18. The Following 2 Users Say Thank You to bjp For This Useful Post:

     Jan_Noack (05-11-2019),  pmokeefe (05-10-2019)

Page 1 of 2 12 LastLast

Similar Threads

  1. Replies: 2
    Last Post: 03-01-2017, 03:08 AM
  2. Replies: 27
    Last Post: 09-24-2015, 02:43 PM
  3. Replies: 10
    Last Post: 07-01-2015, 03:51 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •