Flee at Once ! All is Discovered !

In the last few years, law enforcement has begun to cash in on modern genetic technology in a new and more powerful way, one that works on perps that have been careful enough to avoid being genotyped.

The perp’s DNA sample is matched against a large database of genome sequences, and if a moderately long subsequence is identical-by-descent, it’s clear that the perp and the matching person in the database share fairly recent common ancestry. The degree of sharing shows, approximately, how recent that common ancestry is.

This has been done for some time with close relatives, which share up to 50% of their DNA. But by now, enough people have been genotyped that there’s a pretty good chance that the typical perp ( of European ancestry) has some moderately distant match in the database: say, a 3rd cousin. We can now detect those matches. A few such matches tells us that the unknown perp is one of the descendants of an ancestor a few generations back – so we now only have to examine a very limited pool of potential perps, perhaps a few hundred instead of tens of millions! After excluding individuals of the wrong sex, impossible ages, deceased, etc finding the real killer often becomes quite practical.

In particular, enough people have been genotyped and have uploaded their info to a genetic genealogy site to allow this. I said of European descent: genotyping and interest in genealogy have not as yet been widespread enough among African-Americans to allow this approach , but that will come, probably fairly soon.

What this means, to a pretty good approximation, is that law enforcement is soon going be able to identify the perp in all cases where DNA evidence is available. Decades of unsolved rapes, assaults, and murders will be cleared up – in as little as a couple of years, if we make a serious effort.

Exceptions? DNA from your evil twin, and DNA from truly obscure groups with no representatives in the US. People will write murder mysteries featuring Andaman Islanders -they’ll have to.

We will not find that all or most prominent people have a criminal past ( I think) – but quite a few will, undoubtedly including people you would never have guessed. It will even include people that I never suspected. Successful guys that rarely think about that perfectly understandable mistake that happened ever so many years ago will suddenly find themselves wearing orange jumpsuits. They’ll wish they had converted their holdings into Krugerrands or bitcoin, vamoosed to Chad or Yemen, but it will be too late.

Unless they act now.

This entry was posted in Uncategorized. Bookmark the permalink.

95 Responses to Flee at Once ! All is Discovered !

  1. teageegeepea says:

    There are plenty of murders without DNA evidence. Drive-bys, for instance. So murder mystery writers can remain ignorant of the Andaman Islanders (at worst they’d be more likely to start writing period pieces).

  2. Jim says:

    One of the Sherlock Holmes stories does feature an Andaman Islander perp. The Sign of the Four or some title like that?

  3. P says:

    some background from Graham Coop’s blog on the use of this technique to find the Golden State killer


  4. Frau Katze says:

    I was just reading yesterday about a 56-year old double murder being solved by DNA. They still had minute samples of DNA from one of victims (she was raped). The perp himself was dead. But a case closed.

  5. JAMES SHEARER says:

    “… genotyping and interest in genealogy have not as yet been widespread enough among African-Americans to allow this approach , ..”

    Actually the so called “Grim Sleeper” was identified by such a search. From Wikipedia:

    “Police had found no exact match between DNA found at the crime scenes and any of the profiles in California’s DNA profile database, so they searched the database for stored profiles that demonstrated sufficient similarity to allow police to infer a familial relationship. They found similar DNA belonging to Franklin’s son, Christopher, who had been convicted of a felony weapons charge in 2008. Christopher was too young to have committed the murders, but the familial DNA match led investigators to look at his father, Lonnie, as the likely perpetrator.[43] ..”


    “… In 2004, voters passed Proposition 69, which requires that DNA must be collected from all felons and everyone arrested on some specified charges. ..”

  6. Ilya says:

    I suspect, the barrier of entry into serious crime will become greatly heightened, as a result (for those who wish to avoid being caught, that is). Unabomber level trickery might just become essential.

  7. uhoh says:

    Loved this post! You still got it!

  8. Cat Rationalist says:

    That’s assuming authorities want to.

  9. pyrrhus says:

    With more than 90% of the murders in the US committed by minorities, few of whom are on databases, this technique will be of limited use…There’s also the problem that police forces discriminate against smarter applicants (there are legal decisions approving this), and that defense attorneys would likely be able to shred the chain of inference unless it’s very carefully done….

  10. j says:

    Not long ago, it was possible for a person to “disappear” and show up with another name in the next town. In Japan, it is still possible. The Andamanese would be identified instantly. BTW, Krugerrands is a good idea.

  11. Henry Scrope says:

    I’ve been reading about this for a while, fascinating, probably because it’s new.

    They know who Jack the Ripper was by the way, but there are still TV series being made claiming it was the Duke of Clarence or Queen Victoria’s doctor or some such nonsense.

    • wisewullie says:

      That’s assuming there was a Jack the R.

      • Frau Katze says:

        There were several murders similar to each other in a short time in the same area. He definitely existed.

        I did read something about this. The best suspect was a mentally unstable man. The police of the time even suspected it was him. I don’t remember why they didn’t pursue it.

        • Henry Scrope says:

          Aaron Kosminsky, DNA results from old tissue. You would think it would have been bigger news

          • Jack London says:

            According to wikipedia, the results are highly controversial.

            • Frau Katze says:

              There wouldn’t have been DNA but there is circumstantial evidence, including a complete stop in the killings after the suspect was placed in an institution.

              That’s suggestive for historical purposes but not enough to convict at a trial. Furthermore, restrictions on the mentally ill being convicted was already in force at the time.

              The police may have decided that once he was in the institution they could make sure he stayed there. I think (but I’m not positive) that he didn’t live to old age. They didn’t realize how famous the case would become, with endless speculation on the perp.

              Extra publicity was added by a mysterious letter confessing to the crimes and signed “Jack the Ripper.”

              It is now believed that it was done deliberately by a reporter, looking to gin up interest in the case. He sent the letter to various news agencies. It is unlikely that the average person of that era even know that news agencies existed.

              If he was looking for publicity he succeeded beyond his wildest speculation inasmuch as the name is now part of history.

    • James says:

      And just who was it, pray tell?

  12. jb says:

    Steve Sailer had an interesting thought about the use of DNA databases to solve old crimes:

    The use of DNA from relatives to find murderers and rapists in recent years probably discourages blacks. To whites, the idea of my DNA being used to arrest some 4th cousin for some ghastly crime seems like justice. To blacks, it seem like a way to put a first cousin who is turning his life around in the pen.

    • Woof says:

      A good drinking game is to watch the local news and drink a shot every time a reporter claims a criminal or a murdered criminal ” was trying to turn his life around “. Reason #5489 I stopped watching those idiots.

  13. liquid_crystal_elastomer_man_420 says:

    It’s interesting to think through countermeasures to this – the best path seems to be database pollution, of low-cost emission of chaff. Say one could grow a bag of hair follicles that have random generated DNA to generate a mess of false positives, (or perhaps a deliberate deception).

    Genealogy databases are also vulnerable to corruption. One can imagine a scheme to submit a large volume of samples that would lead to many spurious matches. In some cases it’s likely that the operators of genealogy sites have a financial incentive to look the other way as their data is polluted (like goog scholar with citation counts). Generate a bunch of “missing links” that give everyone a common ancestor with some aristocrat or warrior king. The reason genea sites are a viable business is that they sell validation of everyone’s silly myths about their ancestors. Everyone gets to be a princess! This petty vanity is the key vulnerability.

  14. People will also willingly share information that law enforcement is unable to demand. Not everyone, but enough to make some impact.

  15. jimancona says:

    Razib Khan did a podcast discussing this topic recently: https://razib.substack.com/p/david-mittelman-genomics-for-justice It’s currently subscriber-only, but he removes the paywall a couple weeks after they’re released.

  16. Curious says:

    Greg can we prevail upon you to weigh in with a post on the covid lab-leak versus natural origin debate?

    • HK111 says:

      Same here. But one has to say, the topic is already widely in the open, thanks to Drasticresearch, Nicolas Wade, Steve Quay and many others.

  17. Citizen AllenM says:

    Just wait until people have personal guard drones armed and equipped with tracking implant technology. Just shoot a microscopic rfid chip into the perp, and the next time they get arrested, the chips are scanned, and the guard drone video is uploaded to the prosecution computer.

    Viola, super high tech prosecution. To be followed by a sentence to Running Man!

    Seriously, just have the drone taze them. The future is here, just not evenly distributed- and it will be near impossible in a surveillance society to continue with antisocial behavior.

    The facial recog database industry is just getting started. Now, my personal phone will warn of the approach of an undesirable person, and then it will alert the drone of the security threat.

    I am surprised that drones are not being deployed in Beverly Hills to keep those pesky homeless out of the bushes.

  18. teageegeepea says:

    Off-topic, but Scott Alexander discusses your theory of Jewish IQ in response to Noah Smith’s musing about Jewish achievement not being that “interesting”:

  19. Joseph Marshall says:

    Maybe it will work that way, but I am somewhat skeptical.

  20. JMcG says:

    I love the thought of these dirt merchants wondering if the law is getting ready to knock on the door in the matter of some poor girl they raped and strangled in a field forty years ago. I hope they never get a good night’s sleep again.

  21. Rob says:

    Totally off topic, but on topic for this blog

    Chuck at human varieties did an interesting analysis of IQ and g factor PGS for religion and ethnicities in the US. Human varieties.org does not have it, but Archive.is does.
    Note first chart has PGS on x axis, others are reversed
    Most interesting are ethnicities and denominations far off the regression lines (R^2 from .9 to .97) Like on iq, PGS religion for whites only, Mormons, mainline (I assume) Protestants, Christian, and Nothing in Particular religions have almost identical IQ, but very different PGS (listed above from highest to lowest. Why do the genetics differ so much? What holds Mormons back relative to their common variants (and linkage with rarer stuff)? Why do the nothings have such high IQ, given their genes?

    I can see two possibilities right off the bat
    1) Mormons have more individually rare, ‘private’ alleles depressing IQ
    2) Mormons have a crappy environment for IQ, given their genes

    • Anonymous says:

      What is PGS? Presumably should be eduPGS. I cannot find a definition. I gather it is connected to educational achievement, but can you explain further.

      • Rob says:

        PGS is polygenic score. It/is PGRS for phenotypes that aren’t diseases. It is either for education, but it might be polygenic score for IQ, since in at least one graph it’s called g something ot other.

        • Frau Katze says:

          What’s PGRS? I’ll have to read around on polygenic score. I’ve never heard the term.

        • rankorder says:

          Where do they get their test dataset from?

          The big test of these things is, like, not so much the rank order of groups but within groups.

          I.e. if you score is saying Black doctors are dumber than White mailmen, then it’s probably confounded by ancestry and not a useful tool, even if “the rank order if populations seems right”.

    • Frau Katze says:

      Looking at the second two charts, it seems that Mormons have the highest IQ of the religious folk, just based on the x-axis. But of course I’m still not sure what the y-axis is. But they’re high on it too.

      • Rob says:

        I think you are reading the graph wrong. The y axis is the polygenic score, how many positive/negative alleles they have.

        Mormons have the highest PGS of the 4 religions/ not religions . but approximately equal actual IQ with the other groups. Like, f you you move to the left, to the place that people with that polygenic scores ‘should’ be is much higher IQ than they have. to where their genes predict they should be much higher IQ. Something or other is depressing their IQ.

        I’d really like to see an IQ or educational achievement on West Virginians. What can be done to help them if they have a 95 average IQ than if the brain drain and selection has drastically reduced their IQ.

        • Frau Katze says:

          I’m not understanding what the charts mean because I don’t know what eduPGS means. I entered an earlier comment asking what it means, but I accidentally did not enter it as a reply to you. I was hoping you’d be notified of all comments.

          I looked around but every link to a search on “eduPGS” assumed the reader already knew what it meant.

  22. Frau Katze says:

    What is PGS? It’s too vague a set of letters to find, unless you’re discussing pre-implementation genetic screening?

    • Rob says:

      Ou, Kim sorry. I thought that being a regular, you would know.. a polygenic score for an individual is the phenotype (say IQ score) predicted from his genes, more accurately, the genetic variations, or SNPs actually tested. PRS is the. sum of the effects of common alleles that were associated with that phenotype from a GWAS. It is, in some sense, their “genetic IQ” pRS scores for IQ/educationalcan can account for somethind like 15& of the variance, which is just the R^2 from a regression of PRS vs IQ. For a bunch of people, the graph of PRS vs IQ is not a straight line, but IQ tends to go up with as PRS increases.

      GWAS is genom wide association study. They take a bunch of people, take each ones measure on some trait (phenotype) say from a questionnaire of last grade completed or an IQ score. Then they genotype each person for common SNPs, single nuclear polymorphisms, which are common genetic differences, and see if people with zero, one, or two,copies of that SNP differ from each other on average for For most SNP, the answer will be no, they are basically the same IQ with that all combos of that SNP.of each gene (or intergenic difference). For some SNPs, people with one version of the gene will higher IQ th higher IQ than people with the other version. Like a carton example they test for SNP rsBunchofNumbers with an A Version and G version. The average IQ forbpeople with the three genotypes are AA 99, AG 100, and GG 101, then, if the difference is statistically significant, don’t know how this is calculated), then the G allele raises IQ by one point. They do this for all the genes tested.

      Once they have all the correlated IQ genes, they then sum up all of each person’s calculated IQ from their genes, each person’s sum is their PGS. If t is the best guess of someone’s IQ if all you know is their genome.

      When you plot IQ vs PGS for IQ for a bunch of people, you get a cloud of datapoints with an upward sloping regression line, but it is nowhere near a perfect fit.

      For individual IQ they can predict, have R^2 of .15, 15 percent of variance (variance is a measure of difference from average) but if you take a bunch of people, everyone in the study (or another dataset who’s, say, a Mormon and average their IQ scores and genotypic IQ, the polygenic (risk) scores. you get a group’s average PRS.You get the average IQ of Mormons invthe sample and tithe average you would predict IQ from their combined PRS. Distance from the best fit regression line should be reduced by averaging a bunch of people. When some groups are far from that regression line, that may be interesting

      I hope this helps. Hope I’m not bringing too much coal to Newcastle. Oh, that’s called mansplaining, now, I think?

      • Frau Katze says:

        I still can’t follow your definition completely. All abbreviations like GWAS should be explained.

        This might as well be in Greek: “For some SNPs, people with one version of the gene will higher IQ th higher IQ than people with the other version. Like a carton example they test for SNP rsBunchofNumbers with an A Version and G version”

        My interest in the topic was satisfied by reading Charles Murray’s book. I have no formal education in biology or DNA. My background is physics and math. I know all about statistics and charts.

        I keep telling myself I should educate myself on it, but I don’t seem to have time despite being retired and sheltering strictly at home whilst waiting for my second dose of Pfizer vaccine (Canada is way behind the US. Someone my age would have been fully vaccinated in the US by now.)

        I’m busier than ever after retiring.

      • Rosenmop says:

        I read this book Blueprint, by Nicholas A. Christakis?

        According to this book things such a personality traits and IQ are highly influenced by genes, But each trait is influenced by thousands of SNPs and we just don’t have enough data yet to make meaningful predictions about things such as IQ, or mental illnesses. The author believes that eventually we will have enough data. At least this is what I recall from reading it.

        • Rob says:

          The interesting thing here is not that alleles correlate with traits. The interesting thing here is that groups (with different means) fall far off the regression line for group IQ vs PGS.

  23. LOADED says:

    I have uploaded my DNA to many sites and found a lot of useful information. The best site to use is GenomeLink which gives informative knowledge on all measures of human traits.

  24. j says:

    Flee at once! Darwin’s evolution is out to kill us. COVID has mutated and now it is called Delta!

    • Henry Scrope says:

      It can’t be that bad, the British government is letting 20 or so planes in from Delta every day, with thousands of Deltans on board.

  25. Jaim Klein says:

    “Disappearing” is not a crime of interest for the police in Japan.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s