Genetic Genealogy, Bryan Kohberger, and Parallel Construction in the Idaho 4 Investigation

A legal battle behind closed doors may be the most important and consequential legacy of the identification and prosecution of Bryan Kohberger for the murders of four University of Idaho students in November of 2022.

The State is fighting the defense discovery motion with regard to the use of forensic genetic genealogy during the investigation, claiming it is “outside the scope” of the discovery statute and immaterial to the defense. (And in doing so has confirmed its use.)

I expect the defense to file a more specific motion to compel and to argue vigorously that disclosure is in fact vital to their case. The judge’s ruling may well determine how this case is both prosecuted and defended, with legal reverberations to future cases in which genetic genealogy is used to develop suspects. It certainly will affect how the case is covered, how the the investigation is memorialized in books and documentaries, and may finally lead to the general public understanding that law enforcement narratives often contain omissions and a quasi-legal retro-fitting known as “parallel construction.”

That Bryan Kohberger was identified as a suspect by forensic (or “investigative”) genetic genealogy has been an open secret for months. [The New York Times wrote a full reported article on June 10, a week after I posted this piece.]

The affidavit for Kohberger’s arrest, replete with a gripping narrative of chronological, cause-and-effect policework, is largely a fiction in the way a glowing profile of a young, successful entrepreneur might omit the fact they have a gigantic trust fund. In fact, the investigation was stalled, and would not have been solved without the minuscule amount of touch DNA on the knife sheath Kohberger left near his first victims and genetic genealogy.

The prosecution can and does leave investigative steps out of probable cause affidavits. And PCAs can be written in a way that *implies* a chronology of investigative steps and obscures the chaotic and multi-pronged swirl of simultaneous methods and processing of leads. This is called “parallel construction” — “Parallel construction is a law enforcement process of building a parallel, or separate, evidentiary basis for a criminal investigation in order to conceal how an investigation actually began.” [Wikipedia ] To some extent, parallel construction is a giant lie of omission.

The prosecution likely would prefer not to present witnesses and evidence regarding forensic genealogy at trial, because it adds complexity and is less specific than evidence they subsequently developed. There is also currently a massive behind-the-scenes scrum among law enforcement, the legal community, and the genealogy community about the use of genetic genealogy in active (versus “cold”) cases. The FBI, which has stood up its own genetic genealogy lab, specifically instructs police not to mention its use in arrest affidavits and legal filings. The argument is that it is merely a “tip” or a “lead” or a “technique” that, once the suspect’s actual DNA is gathered and other supporting evidence is amassed, is both irrelevant and a distraction.

More to the point there are moral, legal, and privacy rights ramifications to the increasingly widespread use of ordinary Americans’ genetic material, and the backlash may well hamstring law enforcement going forward. (The moral quagmire and lack of legal boundaries is well exemplified in a recent case where a *rape victim’s DNA* gathered from her forensic exam was used to link her to, and charge her with, an unrelated crime.)

Meanwhile, for the defense, the idea that *this* process of developing suspects is exempt from discovery and cross-examination when all other investigative procedures, processes, tips, leads, personnel, reports etc are not, even if they proved fruitless or are not used in trial by the prosecution — is absurd and smacks of a kind of unconstitutional secrecy treading on the 6th Amendment right to a fair trial and the right to confront evidence.

There are excellent articles and videos on forensic genetic genealogy online, as well as a fascinating, extremely well-done podcast on the first crime solved with the technique [the Bear Brook podcast. Uncovered an unknown serial killer, after which the same genealogist was tapped to identify the Golden State Killer]. But briefly, crime scene DNA that cannot be matched in CODIS, the nationwide law enforcement DNA database, is subjected to a different extractive process and then uploaded to Family Tree DNA and/or GED Match, the two commercial databases where users can opt in to allow law enforcement comparisons. Distant relatives of the person who left the DNA at the crime scene are identified and then a genealogist reverse-engineers a family tree that will include that person. Law enforcement then looks at the likely suspects within that family tree, narrowing by location, age, criminal history, etc. The final step is to directly obtain suspect DNA for a match, which is often done serreptitiously.

********

A close reading of the probable cause affidavit for Kohberger’s arrest reveals a subtext, almost invisible ink, of a parallel, secret investigation.

The key to understanding why it is likely that investigative genetic genealogy first brought Kohberger’s name into the investigation is the white Elantra. And more specifically *when* and *why* the FBI analyst expanded his expert opinion from 2011–2013 models to 2011–2016 models.

Why is the Elantra the key? Two reasons: 1) why did the WSU officers *immediately* search for and find Kohberger’s car in the parking lot, within the hour of finding it in their database? There was reporting that there were 90 white Elantras registered at Idaho State alone. So why did this car get immediate action? 2) at the point where Moscow police asked fellow law enforcement to keep an eye out for white Elantras, it was specifically for the 2011–2013 model, which Kohberger did not have.

Logically, and the defense would have immediately noticed this and circled this part of the affidavit with a big red pen (metaphorically) — Moscow police already had Kohberger’s name.

Another reason to believe investigative genealogy was used is that, although the affidavit is written as though the camera footage of the car’s journey was pieced together contemporaneously, it clearly was not. That is because they had zero footage after he left Moscow heading south. You can see in the affidavit that his path and time of travel immediately after the murders was only determined from cell phone pings after they got the expanded Verizon data on Dec 23, well after they had his name.

Here’s my hypothesis on the sequence of events: They identify Suspect Vehicle 1 via local King Street camera footage and are able to track it coming into the neighborhood, circling around, and leaving at 4:20. They do not get the license plate number; they do notice it has no front plate.

Simultaneously they find the sheath right away and understand this is a key piece of evidence because, unlike searching for the murderer’s blood, which will take a long time given the sheer amount of blood evidence to be processed, the sheath most likely belongs to the murderer and is likely to have his DNA unless he used gloves every single time he handled it from point of acquisition. I also bet that, from decades of experience, the FBI knows that criminals know to wipe down weapons but miss key parts, and such a key part on a sheath is the interior crevice of a sheath snap.

They send the sheath to the Idaho State Police crime lab for testing & a single source of DNA is identified. Single source is *huge.* They fast -track identifying it. It’s not in CODIS, the nationwide DNA database of convicted felons.

Meanwhile the FBI analyst has given his opinion that the video footage shows a white 2011–13 Elantra.

Moscow police decide to try genetic genealogy. The rumor mill keeps mentioning a “specialized lab in Texas” and this is almost certainly Othram. They submit the single profile to the two databases where people have consented to police access. They get a hit on a relative, and I think fairly close relative because of the speed. A genealogist works up a family tree and identifies a set of names of men (the DNA is male) within a certain age range (probably worked up by the behavioral analysis unit). Police use public and law enforcement databases to query those men. [see my July 2 addenda below for the fishy shenanigans behind the scenes.]

Remember that Kohberger was on the WSU criminology Phd website — with a picture. Kohberger also comes up in law enforcement databases for two local traffic stops. Now things are happening fast and simultaneously. They can look at the bodycam footage, they can see his driver’s license, they see his make model and license plate number.

They have found a person who is local, fits the eyewitness description, and drives a white Elantra with Pennsylvania plates, and Pennsylvania does not use front plates. But his Elantra is the wrong year according to the FBI expert. They go back to the expert, get him to revise his opinion. They find a way to get WSU to locate the car without directly asking for his name. And, now that they know where he lives and works they pull WSU and Pullman camera footage to reconstruct the beginning and end of his journey on the morning of November 13th.

They supplement and bolster this with cell phone pings once they receive those from Verizon in December, which also allows a reconstruction of his route south on the following day, and thence actual video footage of Kohberger in a grocery store. I’d be very interested to read the subpoena for phone records to Verizon; what probable cause was given?

The whole question of when and how to approach him was probably hotly debated.

Now they have probable cause, but he has already left on his road trip to Pennsylvania. I don’t think he was already being tracked because remember about that time in an interview Chief Fry sort of slipped up and said “we don’t know where the suspect is,” then quickly added “we don’t know who he is.” They do get a hit from a license plate reader in Colorado and then two visual confirmation traffic stops in Indiana. It’s clear he’s headed home for the holidays.

They still need confirmation of his actual DNA matching the sheath sample. There has been at least one case where a suspect identified through genetic genealogy, and who was in the right place and the right time, and was the right age, was not a DNA match to the evidence and was completely not involved in the crime. Coincidences do happen. They likely had been trying to get discarded DNA from him through early December with no luck, and had to settle for his family’s trash, hoping for his but instead getting his father’s, which is sufficiently dispositive. After the arrest they promptly collect his DNA directly to match to the knife sheath.

The identification of the DNA from the Kohberger family trash as belonging to the father of the person whose DNA was left on the knife sheath took just one day. That’s because it is a commonplace process called “familial DNA” available to any decent crime lab. However, multiple writers and true-crime YouTubers have mistaken this, final step prior to Kohberger’s arrest, as “genetic genealogy”, and confusion reigns. I’d call out most egregiously the author Howard Blum, who is writing a book on the case, a book destined to read like pulp law enforcement fanfic and to bear absolutely no resemblance to the actual investigation, the actual legal process, or any person living or dead.

*********

Anyhow, for the defense, they can probe the causal investigative sequence to reveal that Kohberger was the target early on, around November 29th. [the New York Times places this at December 19th.] They can drill down on when and why the model of car identification was expanded, and accuse investigators of retrofitting evidence to their theory, a type of tunnel vision. They can ask why the call was put out for 2011–13 Elantras even after Kohberger was identified; was that simply cover for the fact they had already closed other investigative avenues?

I absolutely want the judge to rule in favor of the defense, and for the prosecution to fully disclose the who what when and where of the genetic genealogy piece of the investigation. I’m extremely interested in *actual*, granular investigations as opposed to heroic narratives or glib parallel constructions. I’m interested in fact-based coverage and discussions. I think the more charged an event or issue, the more vital it is to adhere to the truth.

Much more importantly, law enforcement has used parallel construction to obscure the actual sources of critical evidence from view for decades. This practice denies defendants and defense attorneys multiple avenues to challenge the admissibility of evidence, the credibility of informants, and the legality of hidden aspects of the investigation.

I have experienced this, on a small scale, directly myself. A few years ago I provided a tip to law enforcement that directly initiated an investigation leading to an arrest. In the two media reports of this arrest I was bemused to read the lead investigator claiming my tip as a result of her own effort. I believe this was done to protect me, but to see that I was entirely erased from information that I and only I had handed them, in my exact words, was eye-opening. It happens all the time.

July 2 Addenda

The defense did indeed file a spicy motion accusing the prosecution of hiding the football. The motion adroitly pointed out that the probable cause affidavit for Kohberger’s arrest omitted all mention of investigative genetic genealogy, while multiple investigative steps and search warrants were clearly subsequent to its use and reliant upon it. There is a vague hint of possible fruit-of-the-poisonous-tree issues.

More ominous is the expanded motion to suppress by the state which casually reveals that a) while the knife sheath DNA *was* sent to a private lab to generate the SNP profile and begin to construct a family tree, law enforcement decided to instead assign the work to the FBI; and b) the FBI uploaded the DNA profile to genealogy database(s), received matches of relatives, constructed a family tree, identified Kohberger and gave that “tip” to Idaho law enforcement — and then deleted the profile from the database and did not memorialize any of their work nor create a report.

The state seeks to suppress the family tree itself, claiming it violates the privacy of the “DNA informants” (a new legal category, which, if ratified by this judge, will have major consequences going forward). They also seek to exclude from discovery the identity of the private lab and all its work product.

The extreme secrecy around this pivotal investigative work seems to me to indicate that something fishy is going on. And my hypothesis is that the private lab found very distant matches and began a family tree that was going to take a lot of painstaking, time-consuming work. The FBI then appropriated the DNA profile and did the work in-house and black box. And what I think they did was access off-limits profiles in multiple databases. I think the FBI submitted the sample as if they were a private citizen using the product. I also am beginning to think, now that it was revealed (in late August 2023) that a genetic genealogy witness for the defense received an ominous visit from the FBI subsequent to testifying at a hearing, that something even more shady may be afoot, such as the FBI secretly creating and using its own SNP database. (Othram is doing the same; creating an in-house repository of SNP profiles. That is not conjecture; they have a website soliciting DNA samples from the public.) Any of these possibilities exponentially increases the likelihood of getting a close relative match, and that is what I believe happened. Then the FBI deleted their work, closed the portal, and hid behind the legal constructs of “tip” and “informant.”

The very reason that 23 and Me and Ancestry.com, the biggest genealogical services, closed their databases to law enforcement in the first place was because of abuses and lack of transparency. In a case I link below a completely innocent guy was brought in for custodial interrogation, and his DNA taken by warrant, after he was identified by genetic genealogy as a possible suspect. His father had submitted a DNA sample to a family tree project run by his own church, which, unbeknownst to him, ended up uploaded to one of the major genealogy databases. That service faced media blowback and prospective class-action lawsuits, and promptly adopted a policy of nondisclosure to law enforcement.

If I’m right in my hypothesis (and I think Kohberger’s defense is also thinking along these lines), this thing is going to blow up on the prosecution and possibly change the legal landscape to allow defense discovery of investigative genetic genealogy going forward. We might even get one of those rare open court moments where a defense lawyer asks an FBI agent on the stand “which service did you submit the sample to?” and they have to admit, under oath, that they circumvented the rules and possibly the law.

— — — — — — — — — — — — — — — —

SOURCES AND FURTHER READING

Inside the Hunt for the Idaho Killer (June 10 2023) NY Times
https://www.nytimes.com/2023/06/10/us/idaho-university-murder-investigation.html

How Police Actually Cracked the Idaho Killings Case (Jan 10, 2023)
https://slate.com/technology/2023/01/bryan-kohberger-university-idaho-murders-forensic-genealogy.html

Kohberger Probable Cause Affidavit
https://www.documentcloud.org/documents/23564645-kohberger-moscow-pd-probable-cause-affidavit

Prosecution Notice to File a Motion in Limine Excluding GG From Discovery (May 12, 2023)
https://bloximages.newyork1.vip.townnews.com/kxly.com/content/tncms/assets/v3/editorial/7/42/7429e89a-f383-11ed-90b7-8ff43e7088f7/6462d32d5e68f.file.pdf

Legal Analysis of Prosecution Argument to Exclude IGG from Discovery
https://www.youtube.com/watch?v=CD8Z25NIB9g

CeCe Moore on CNN — Noted genetic genealogist Moore gave multiple interviews to various outlets both prior to and immediately after Kohberger’s arrest indicating she knew genetic genealogy had been used
https://www.youtube.com/watch?v=0IXdMiHsYYs

Genetic Genealogy Leads to Interrogation of Innocent Man
https://www.cbsnews.com/news/angie-dodge-murder-how-a-discarded-cigarette-led-to-an-arrest-in-idaho-falls-teens-cold-case/

Bear Brook Podcast — utterly riveting, beautifully presented reporting on how genetic genealogy accidentally revealed a prolific serial killer
https://www.bearbrookpodcast.com/season-one