To catch a predator: familial DNA

By Razib Khan | July 8, 2010 10:51 pm

I already blogged this general issue, but the ‘grim sleeper’ murderer was caught because of a match of old samples with those of us his son. If I had to bet money I think this sort of result (California and Colorado are the two American states which have a system in place to allow for this) is going to allow for a push toward more widespread usage of the technique. It may be that we need to stop talking about privacy as if we can put off the inevitable future, and start talking about accuracy and precision with the data that is going to be easily available to authorities. By the way, I found this objection somewhat strange:

“I can imagine lots of African-American families would think it is not fair to put a disproportionate number of black families under permanent genetic surveillance,” said Jeffrey Rosen, a law professor at George Washington University who has written about this issue.

A disproportionate number of black families have relatives incarcerated. The American public does not seem particular worried about that. As I noted before, criminal behavior is not randomly distributed across families. Rather, there are distinct clusters, so familial genetic data is going to be more efficacious than you would expect if the commission of crime consisted of a sequence of independent events.

I have to add that worries about this technology strike me as a bit rich, in light of the fact that methods which are proven to be highly subjective and often inaccurate, such as fingerprinting and eyewitness identification, are accepted in the criminal justice system. I worry about what the state could do with DNA data if the state became malevolent, but despite its flaws it seems to me far preferable as a means of assessing evidence than some of the “tried & true” techniques. So let’s keep some perspective.

  • Pingback: Tweets that mention To catch a predator: familial DNA | Gene Expression | Discover Magazine --

  • Zora

    It is true that a high percentage of convicted criminals come from minority backgrounds. However, they disproportionately victimize their relatives and neighbors. It could be argued that the use of familial DNA evidence will actually HELP minorities, by reducing the crime rate in minority neighborhoods.

    I don’t see why we shouldn’t worry about the victims just as much as we worry about the perps.

  • Sandgroper

    Sure, but I can see how the argument might run – ‘disproportionate’ rates of ‘genetic surveillance’ of some population subgroups will result in ‘disproportionately’ high rates of criminal conviction among those subgroups, which will result in even more disproportionate rates of genetic surveillance etc.

    It’s predictable that it is being cast as a form of racial profiling.

    However, I see an important difference – with due attention to accuracy and precision, DNA data provide protection against wrongful conviction. So it becomes a very odd argument.

  • Georg

    [i have removed a comment which consisted purely of commenter incredulity, since i have approved previous comments from this individual i’ll give this commenter another chance if they want to say something which elaborates on their genius insights. otherwise, nope. -rk]

  • bioIgnoramus

    “criminal behavior is not randomly distributed across families”: I understand that the same is true of terrorism – or at least that it was for Irish terrorists Maybe it’s true only for terrorists whom the US supported, but that seems implausible.

  • Divalent

    “… lots of African-American families would think it is not fair to put a disproportionate number of black families under permanent genetic surveillance.”

    I think the issue of accuracy and reliability of the DNA information is key. It is the case now that, due to a disproportionate arrest rate, a disproportionate number of AA people have also been put under fingerprint and mug shot surveillance. (Does any one wring their hands about that?)

    Of the 3 (mug, fingerprint, DNA) I’d be most worried about mug shots, since the potential for inaccuarately being tied to a crime I wasn’t actually involved in are a quite a bit higher. (Is anyone complaining about mug shots?)

  • miko

    Razib, I have to say your trust in the state’s curation and application of DNA data for the common good is almost…. Canadian.

  • Caledonian

    This is one of the issues where there are no intelligent arguments against – so people end up spending a lot of time arguing against stupid arguments.

    Partial matches are something that arise naturally from searches. The authorities already have the information – they would have to deliberately blind themselves to the data they possess, or deliberately fail to draw the obvious conclusions from their searches. If the DNA found at a crime is similar but not identical to that of a registered criminal, looking at relatives of that criminal is a no-brainer.

    Just as if the description of a suspect closely resembles but doesn’t match that of a known criminal, and his family members look like him… except that DNA is far, far more specific than appearance.

  • Razib Khan

    Razib, I have to say your trust in the state’s curation and application of DNA data for the common good is almost…. Canadian.

    dude, that was low! though to be fair, i don’t have that much trust…but we give the state a monopoly on criminal justice already, and this is pretty much inevitable i think.

  • Martin

    An interesting aside: Adrian Lamo is the “hacker” who ratted on Bradley Manning, the military intelligence specialist now accused of leaking the “Collateral Murder” video and supposedly a quarter million State Department memos that have yet to be released, and which Wikileaks denies having.

    Lamo was convicted of breaking into the computer systems of several prominent companies back in 2002/3. While he was on federal probation, he was required to give a blood sample so that his DNA could be added to a national offender database. He refused on religious grounds (because the Bible says something about spilling blood being a sin). He brought fingernail clippings to the probation office instead. The probation office only accepted blood samples. He went to court over this and faced the possibility of prison time for probation violation, but eventually won the case and was allowed to give a cheek swab instead.

    It’s interesting to me because it speaks to the issues of DNA collection and religious exemptions on legal requirements.

  • Nick Patterson

    I’m a geneticist, and the issues here are more complex than they seem.
    For instance, samples I collect for research have to be made available to the
    authorities if they re investigating a crime. Thus participating in a research
    study, can in principle lead to your DNA being used to convict you, or a close
    family member.

    This can (and does) reduce the numbers of people willing to contribute to research.
    It’s unclear to me what’s right here. But for example, if you commit a speeding violation
    are you happy for your DNA to then be in police files forever?

  • bioIgnoramus

    When I went to hospital recently, they swabbed my mouth and nose to check for infections. I suppose they could have recorded my DNA had they wanted to But since they also took lots of blood samples, they’d have been spoiled for choice.

  • Razib Khan

    the issue with hospitals is a major future point. theoretically the cops could just get a court order to look into those dbs.

  • Pingback: Police Nabbed Serial Killer Suspect by Stumbling on His Son’s DNA | JetLib News()

  • Bootlegger

    “criminal behavior is not randomly distributed across families”

    Yes it is. If you include all types of proscribed behavior you’ll find that 95% of people break the law. The labeling of behavior as criminal, however, is not randomly distributed across families. Some “families” are far more likely to wind up in prison than others even for the exact same behavior. Lower SES people will be more likely to be in the database than high SES people and if the database is used to find more criminals it will perpetuate that disparity.

  • Caledonian

    But for example, if you commit a speeding violation
    are you happy for your DNA to then be in police files forever?

    How is that different from the taking of fingerprints? Currently fingerprints are taken whenever people are arrested for any reason.

    There’s a police file that hangs around forever if you, for example, commit a speeding violation. I’m not happy about that, but I see no particular reason to be more unhappy about that than the file having DNA information in it.

    The real concern is one that isn’t being brought up here: fingerprints tell us nothing else about the person they’re from, while DNA might potentially tell us far more.

  • Nick Patterson

    I agree with #16 that fingerprints on file is a related issue, but there
    are differences. One is that your DNA is shared by your relatives, so
    could be used to convict one of your children, even after you are dead…

  • Sandgroper

    So what? You think your kids should be exempt from prosecution if they commit violent crime, just because they’re your kids? I’d sooner feel that my kid is better protected from violent crime. She has the choice not to commit violent crime.

    It gets better. My office was burgled once, and the police took a full set of my finger prints, my secretary’s prints, and the prints of my colleague who was covering my job while I was on holiday, so they could exclude us from the prints they found in my office. They were careful to explain why they had to do it, but then said “But of course, once they’re on the record…” …they stay there.

    You don’t need to commit any kind of violation to get your prints on the record. You just need to be the victim of a crime, or the legitimate associate of a victim.

    Given that, I’ll take the DNA data in preference, provided there are adequate safeguards on accuracy.

  • Biologist

    I didn’t see a clear enough description of what’s actually happening in cases of familiar searching — perhaps it’s obvious — but I thought it was worth spelling out explicitly.

    1. A murder occurs and a DNA sample is obtained as evidence from the environment of the crime (e.g. a blood stain). The unknown person that is the source of the DNA is suspected of committing the murder.

    2. That environmental sample is genotyped using 13 multi-alleleic markers. There are no direct suspects or all direct suspects are ruled out by failure to match the profile of those markers, so a search of the CODIS database is performed. However, no perfect 26-point matches are found (13 variants and two alleles per variant). Thus there are no direct DNA matches against people who for various reasons have had their data added to CODIS — largely people convicted of crimes.

    3. The search criteria is loosened and a “familial search” is performed. This can be done in a more or less statistically robust way, depending on the type of familiar relationship being examined. In a father-son match, the son will match at least one of the father’s alleles at each of the 13 loci. This kind of search is loose enough, for example, you get 200 hits again the environmental sample’s profile that could be a first degree relative of the. Note that none of these individuals are suspects in the case — they are not a direct match. However, their relatives are now potential suspects for the crime. The familial search thus produces a very long list of suspects.

    4. At this point regular police work could resume and the list of suspects could be narrowed — for example by the life history of the person for whom the partial match was found (e.g. did they have a living 1st degree relative at the time of the murder who was of the right age to commit the crime and also had opportunity for geographic reasons). If it can be narrowed to a handful of individuals — relatives of individuals who are in the DNA database but ostensibly none of whom are themselves in the DNA database — then an effort can be undertaken to obtain a new DNA sample from the one or more suspects to test whether they actually match the original environmental sample from the crime scene.

    5. It sounds as if the matching DNA sample in this case was acquired surreptitiously — an entirely different legal/ethic issue. But the goal of that effort is to confirm or refute a match of the new suspects against the original profile. In this case, a positive match was obtained from the father of an individual who’s profile was in the DNA database. This is very strong statistical evidence that the father was the source of the DNA found at the crime scene.

    In summary, familial searching identifies a pool of potential suspects for a crime based on the resemblance of one of their relatives to the DNA found at a crime scene. It does not by itself establish which of those suspects — if any — was the criminal. But in conjunction with normal police work, it can lead to a short list of high probability suspects for which it might be possible to establish a more definitive judgment based on direct DNA testing.

    The process seems to be sufficiently labor intensive that it would likely be reserved for the most serious crimes.

  • Razib Khan

    you also listen to an interview with experts about the method in reference to this particular case:

  • Biologist

    Setting aside the racial inequalities in who is found in databases — consider that in this case the victims of the crimes were themselves black, iirc — the real downside of familial searching is that it seems to violate the spirit of the original laws which set out who should be included in DNA databases.

    Eventually, a de facto universal DNA database will probably be assembled through direct or indirect methods. Just picking one example, a court order will be able to open a personal medical record given probable cause, and that record will eventually contain a full DNA sequence.

    Although politically impractical, a proactive effort to setup a a universal database in which:
    (1) only markers that are phenotypically of little informativeness are used
    (2) the original DNA sample is destroyed
    (3) strict access rules and searching criteria are established
    would probably be the optimal solution.

  • Georg

    I recommend You read some book on daktyloskopy.
    Blaming fingerprinting to be “subjective” still is
    nonsense, You cant change this fact by erasing
    a comment.
    Dctyloscopy is, compared to genetics, maybe less
    sensitive, but never it is subjective.
    I think You do not know how it works, maybe its not
    enough “hightech” .
    Meanwhile some other blogger here has copied this
    sentence. :=(

  • Neil

    Like the Biologist @19, I was expecting a key issue to be the number of false leads generated.

    However, ScienceInsider has a subsequent piece on this

    California allows familial DNA searches only for violent crimes in which the perpetrator is still believed to be a danger to society. Sims and Myers say they have run 10 searches so far. The first nine came up empty, including a 2008 search with DNA evidence from the Grim Sleeper crime scene. “We did not find anybody in the database who we thought was a potential relative,” says Sims.

    However, a second search in April 2010 did turn up a potential match: a young man named Christopher Franklin who was convicted last year on a felony weapons charge. The DNA search, along with the dates of the murders cast suspicion on Christopher Franklin’s father.

    This from the DNA profiles of “1.3 million convicted felons”.


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Gene Expression

This blog is about evolution, genetics, genomics and their interstices. Please beware that comments are aggressively moderated. Uncivil or churlish comments will likely get you banned immediately, so make any contribution count!

About Razib Khan

I have degrees in biology and biochemistry, a passion for genetics, history, and philosophy, and shrimp is my favorite food. In relation to nationality I'm a American Northwesterner, in politics I'm a reactionary, and as for religion I have none (I'm an atheist). If you want to know more, see the links at


See More


RSS Razib’s Pinboard

Edifying books

Collapse bottom bar