“Lake Stickney John Doe” spent almost a decade underwater before surfacing in June 1994 — a pair of fishermen found him among a patch of lily pads in Snohomish County, Washington.
The body was decomposed to the point of corpse wax — a blue-gray clay-like coating on the skin, the body waterlogged and bloated with bacterial detritus. An autopsy revealed only a few details: He had been shot in the head, likely dumped in the late 1980s and was somewhere between 25 and 35 years old.
Discovering his identity would be an uphill, if not impossible, feat. One-fifth of a nanogram — fewer than 20 human cells worth of DNA, all of it incomplete and contaminated, were recovered from the body. This, compared to a traditional cheek swab which generates between 750 and 1,000 nanograms of clean, complete DNA. The case stayed cold for over 26 years.
This summer, as lakes dry up due to climate-change induced drought, there have been an alarming number of similar John and Jane Does found — their remaining genetic material compromised after being submerged years in lakewater. At Lake Mead, four bodies have been discovered in the reservoir since May, the latest turning up on Aug. 6.
A new technology that recovers “itty-bitty bits” of DNA
Large lakes — Lake Mead, Lake Tahoe and Lake Michigan among them — have long been mythicized as convenient dumping grounds for mob-era gangsters and killers. And until a new-developing field of science called genetic genealogy broke the cold-case scene wide open, they were, more often than not, unsolvable.
The growing sector of genetic genealogy combines advances in two distinct scientific fields — DNA sequencing and genealogy. Scientists are able to recover trace amounts of DNA to build a genetic profile and then infer familial relationships by searching databases of DNA profiles from people who have paid to have their genetic material sequenced. Though only a decade old, the method has helped solve cases where law enforcement had DNA samples, but nothing to compare them to.
“The real leap that takes us into the future is perfecting the methods to take even the crappiest, most challenging, degraded DNA and still being able to extract insight from it,” said David Mittelman, the founder and CEO of Othram, a forensic sequencing lab that uses DNA sequencing techniques to “identify victims, find missing persons and reveal perpetrators of crimes.”
Another chance at Lake Stickney John Doe
In 2020, after failed, more traditional attempts at DNA sequencing with other labs, the “Lake Stickney John Doe” cold case and DNA sample came across Mittelman’s desk as a last resort. Using the new method, his team was able to sort between human and nonhuman DNA, distinguish the broken sequences from those that remained intact and repair the broken ones using enzymes.
It’s not a one-person job, said Mittelman. “We’ve got computer programmers that have built models for how to read the data, chemists and lab technicians that work to iterate the lab process,” he said. “There’s a lot of different minds working on this.”
Mittelman and the Othram team were able to reconstruct a DNA profile for the man. But while he had a genetic identity, he still lacked a personal one. Who he might be related to — information that could ultimately reveal his name — was still unknown.
Once a DNA profile is complete, the next step is building out a family tree
The use of DNA tests to determine family history became possible around the turn of the century, as commercial DNA labs — the first being Family Tree DNA — began offering testing kits to the general public. Anyone could submit a cheek swab and their DNA profile would be sequenced and stored in the company’s online database.
There were limits to these initial tests — the ability to trace back one’s paternal side was completely independent from tracing their maternal side. But citizen scientists and genetic enthusiasts gravitated to the practice, said CeCe Moore, the chief genetic genealogist for Parabon NanoLabs, a DNA technology company headquartered in Virginia.
The big breakthrough came in 2009, when 23andMe launched their “Relative Finder” tool. The company was able to collect autosomal DNA — that which is 50 percent from the mother, 50 percent from the father — and could build out extended, complete family trees. Now entire families could submit their DNA and build back their lineages for generations.
“So some of us — me, certainly — jumped in with both feet,” Moore said. “I tested 40 of my family members and started a blog. This was the first time family data was looked at to any large degree. Because there was no such thing as a professional genetic genealogist at that point, I became known as an expert in a field that had no experts.”
Over time, Moore and other grassroots citizen scientists turned their focus outward, using their expertise interpreting and understanding DNA to help people find relatives, working backward to build out their family tree.
The technique is reliant on the existing DNA profiles available in online databases. Genetic genealogists cross-reference an unknown persons’ DNA — those that have been restructured by scientists — with what currently exists and look for matches.
A 1 percent match is usually indicative of third cousins and around 3 percent meant second cousins. Any big hit like this reveals a common ancestor and makes a family tree more complete. Because the identities of those in DNA databases are known, this greatly narrows down who the unknown person could be.
Despite a track record of cracking cold cases, there remain challenges
The field gained fame in 2018 when genetic genealogy was used to identify the Golden State Killer. Many genetic genealogists, including Moore, began including cold cases in their analyses. To date, she’s helped solve over 200.
But the field is not without its challenges, and its reliance on databases is perhaps the largest. About 40 million people have submitted their DNA across all the largest databases — Ancestry.com, 23andMe, MyHeritage — but only Family Tree DNA and GEDmatch allow access for work with law enforcement: about 2.5 million unique samples combined.
And the databases are not representative of society. Most of the people who have willingly submitted their DNA are of European ancestry, Moore said. Other ethnicities are underrepresented, a remnant of mistrust of medicine, law enforcement or both. It is also near-impossible to build a complete family tree of those descended from enslaved people, using DNA alone, because so many names and records have been destroyed.
But confirming identities remains genetic genealogy’s strongest suit
“Lake Stickney John Doe” was one of the identifiable ones. After Mittelman’s team generated his DNA profile, Orthram’s genetic genealogists found a close familial match, which, using public records and traditionally investigative detective work was enough to identify the man as Rodney Johnson.
The future technical challenges for genetic genealogy are its scalability. “The true innovation comes from being able to extend the technology to every case, not just the ones that have good DNA,” Mittelman said. “You want to be able to make sense of any case — someone dumped in Lake Mead, or Lake Stickney, or in a septic tank. You want to be building technology that is inclusive of all cases, not just the easy ones.”
And building family trees in more efficient, inclusive ways will require larger DNA databases and access to more genomes. “Providing answers is really important for those families,” Moore said. “Everyone deserves an identity.”
Thanks to Alicia Benjamin for copy editing this article.