Nobody Was Going To Solve These Cold Cases. Then Came The DNA Crime Solvers.

A woman in a ditch. A baby in a parking lot. The DNA Doe Project has the ability to identify cold case victims — so why don’t they take on every case?

Their first dilemma was the frozen baby.

In January 1988, police in Meriden, Connecticut, discovered the body of an infant in a parking lot. Wrapped in a pink blanket, the blonde, blue-eyed, seven-and-a-half-pound boy reportedly still had his umbilical cord attached. The case tore through investigators’ uniform-hardened hearts in a way few others did. They named the baby David Paul. They buried him under a stone that read “Beloved Little Man” and held annual memorials in his honor.

Nearly 30 years later, in California, two women had to decide whether to finally identify the dead newborn. Margaret Press and Colleen Fitzpatrick possessed the specialized knowledge and technical skills to help these detectives solve their mystery. But should they?

Press and Fitzpatrick are the founders of the nonprofit DNA Doe Project, a network of genealogists volunteering their time to help law enforcement agencies solve cold cases. They deploy the same technique — genetic genealogy, or using DNA to fill in the blanks on a family tree — that led to the April 2018 arrest of California’s suspected Golden State Killer, and then to the capture of cold-case perps around the country, seemingly every other week.

The DNA Doe Project has a different kind of mission. Its volunteers are devoted to the dead, rather than the living. They don’t hunt elusive suspects; they put names to unidentified bodies. So far, they’ve gone public with identifications in three tough-to-crack cases: a young woman found strangled in an Ohio ditch in 1981; a man who checked into a Washington motel in 2001 under an alias, then killed himself; and an Ohio man whose 2002 suicide revealed he’d been living with a stolen identity. Ten more cases are in the pipeline as inquiries trickle in from agencies across the country. The demand is daunting, with coroners and medical examiners reporting more than 12,200 unknown corpses in the US today. But for Press and Fitzpatrick, every DNA file in their charge has become a story, and a mystery only they may be able to solve — not just who the deceased really are, but why no one came looking for them or even knew they were dead.

“You know they had a mother. They had a father. They had Christmas presents. They had a birthday party. They bought clothes at Kmart. They had a life just like the next guy driving down the street,” Fitzpatrick said. “You have to wonder where they came from. ... How come they’re in a ditch? How come somebody stabbed them? How did they get there?”

Yet they didn’t start out with this sentimentality. When they teamed up in early 2017, Press and Fitzpatrick focused on John and Jane Does for one reason: avoiding controversy. Many people knew the methodology of using familial DNA to catch suspected criminals was possible, but there were concerns over privacy and consent. Individuals who uploaded their own DNA to public databases didn’t anticipate cops using their profiles to hunt criminals, let alone criminals they might have seen at their family’s Thanksgiving dinner. Press and Fitzpatrick thought Doe cases might be more palatable and sympathetic: groundbreaking but not boat-rocking.

“How come they're in a ditch? How come somebody stabbed them? How did they get there?”

That all changed with the splashy apprehension of Golden State Killer suspect Joseph James DeAngelo, the alleged rapist and murderer who terrorized swaths of California from the mid-’70s to mid-’80s. DeAngelo was arrested after investigators sought the help of genetic genealogists, who used the public database GEDmatch to develop a family tree based on DNA recovered from a 1980 rape kit. The result was chaos in the world of cold-case cracking, likely the biggest development since Congress authorized CODIS, a national database of DNA collected from crime scenes and from accused and convicted criminals. But most importantly to Press and Fitzpatrick, the public wasn’t outraged by the technique, as they had feared.

A “revolution” commenced, said Fitzpatrick. Over the last five months, genetic genealogy has become the wild west — made up largely of citizen scientists, amateur sleuths, hobbyists, and hobbyists turned professionals, many scrambling to lend their puzzle-solving skills to law enforcement, who are in turn scrambling to understand this new and complicated way of solving old crimes.

Humming underneath the field’s frenetic excitement is trepidation: What happens if someone attempting this technique messes up? When genetic genealogy is tested in court for the first time, will it hold up to a defense attorney or jury’s scrutiny? There’s little regulation or oversight involved when a professional detective sends another human’s DNA to these at-home detectives. Ethical lines are being drawn in real time, and Press and Fitzpatrick have found themselves with pens in their hands.

“It’s the big bang, and it’s exploding in all directions.”

“All these considerations now are happening too fast for us to keep up with it: the legal considerations, how will it play out in court, the ethical considerations, the forensic considerations, the genetic considerations,” Press said. “It’s the big bang, and it’s exploding in all directions.”

Which brings us back to baby David Paul.

Press and Fitzpatrick took the case; identifying him was supposed to be their first victory. But while they waited for their lab to sequence the infant’s DNA, the women did some soul-searching. They realized their task wasn’t really about naming a nameless baby. It was about identifying the mother who left him to freeze to death, so that she could potentially be prosecuted. They were in a bind: Law enforcement needed them, and they needed law enforcement, but how could they fulfill their mission without crossing their own line?

Fitzpatrick warned me before we met at her suburban townhome in Orange County, California, in mid-August: “Keep in mind ... we don’t have a lot of whiz-bang equipment to do a laser light show. We are two little old ladies who work on their computers all the time.” Press also works out of a home office, in a bungalow in a quaint Sonoma County city 450 miles north of Fitzpatrick’s. The two women speak on the phone every day and see each other a few times a year. They first met online, and today that’s their headquarters and playpen — where all their work and the bulk of their interactions occur.

“Pretty amazing that the two of us can change the world from our homes!” Fitzpatrick wrote in an email.

Fitzpatrick is an elvish 63-year-old, whip-smart and funny and prone to describing her work in this grandiose way. But her pride has been earned. She’s put in the hours, thousands of them, pecking away at the keyboard that she plugs into her stocky black laptop. She works late at night and into the early hours of the morning. During the day, she takes calls from agencies. She warned me about this too, and sure enough, on the afternoon we met, the NYPD rang. She took the call out of earshot, pacing around her small kitchen, stopping to write things down in a black-and-white composition notebook.

But there was a time when Fitzpatrick couldn’t even get law enforcement officials to take her calls or meetings. When they finally began agreeing to see her presentation, they’d stare at her blankly. Afterward, she’d ask for their business cards, and they’d decline.

“The excitement that law enforcement have about genetic genealogy, that wasn’t there before” — that is, before this year, before the Golden State Killer arrest, before the big bang. “Now they get it. It’s always been exciting. It’s always been useful. But now they get it,” she said.

“The excitement that law enforcement have about genetic genealogy, that wasn’t there before. Now they get it.”

While genealogists have only recently garnered attention for working with law enforcement, Fitzpatrick has been helping cops solve crimes since at least 2011. Her methodology was different then. Instead of using today’s very thorough autosomal DNA tests — which would take a few more years to become affordable and popular — Fitzpatrick was using the Y-DNA testing method, which only worked on men. (Y-DNA refers to the Y chromosome, passed from father to son.) The technique was popular among adopted men looking for their birth fathers; with the help of genealogists, they could trace their paternal lineage by searching public Y-DNA databases for men with matching genetic markers. But their results would be limited to potential surnames like Smith, Johnson, and Jones.

Fitzpatrick’s first police case was an attempt to identify the killer of 16-year-old cheerleader Sarah Yarborough, who was strangled near her high school campus in Federal Way, Washington, in 1991. (The case remains unsolved.) Using Y-DNA, Fitzpatrick was able to give Washington authorities the Yarborough murderer’s possible last name: Fuller. The police went to the media with her findings. And her peers went ballistic.

Fitzpatrick had stepped on an ethical landmine that she didn’t realize was there. Though she’d only been working as a full-time genealogist for a few years by then, she’d gained some renown for solving sensational mysteries, like identifying the unknown child who died on the Titanic, and coining the term “forensic genealogy.” Fitzpatrick — a nuclear physics PhD who once worked with the likes of NASA and the Department of Defense — wanted genealogy to be more scientific-minded, with higher proof standards, so that it could be used in legal contexts beyond the collection of family history. And she wanted to work with law enforcement, despite her peers’ protests about government overreach and their fear that people would stop taking genealogical DNA tests if they thought their privacy was being violated.

For a time, Fitzpatrick’s then-radical vision made her unpopular; when news broke of the Yarborough case, she got hate mail and was axed from genealogy listservs, she said. She began to feel uncomfortable at conferences, aware that her reputation preceded her.

Which is why last year, when she and Press founded the DNA Doe Project, she knew it was wise to stick to cases involving dead victims or unidentified remains if they were going to use this new GEDmatch technique, which was more precise than Y-DNA testing but also more potentially invasive. GEDmatch could provide users (including police) a list of names and email addresses of living relatives, rather than just a possible surname dating back a dozen generations.

The big companies (, 23andMe) still refused to let cops raid their data, or let users upload anyone else’s tests but their own, but GEDmatch had no such policies. It was “a desert island where all the genealogists can go play in the sand,” Fitzpatrick said. It was also much more difficult to use.

Because GEDmatch’s database was a tenth of the size of, users weren’t finding their uncles or half brothers on it; they were finding their third and fourth cousins, making it challenging to build a family tree. And, for the DNA Doe Project’s purposes, it was expensive, about $1,500 for a sequencing lab to take blood or bone from an unidentified body and churn out a digital DNA file, then for a biology algorithm whiz to take that massive file (with its 3 billion letters of the human genome) and make it GEDmatch-compatible (reducing it to about 600,000 letters).

But to Press and Fitzpatrick, the challenge and costs were worth it. They thought if they could prove their method worked, they could, as Fitzpatrick said, change the world. They weren’t thinking then about how taking on Doe cases might affect them personally.

When she identified suspects, Fitzpatrick felt somewhat removed from the process. She did her work from behind her laptop, then police took her lead and ran with it. “But the Does,” she said, “sometimes you get mixed up a little bit with the family. The family needs to know what happened. We try to keep our distance, but they turn to us because they can’t turn to the policeman and cry.” The genealogists’ involvement became more intimate. That felt good, Fitzpatrick said. “With a suspect, you don’t have that warm feeling.”

This became apparent in the DNA Doe Project’s first announced identification: Marcia King, or “Buckskin Girl,” the 21-year-old murder victim discovered in an Ohio ditch in 1981 wearing a buckskin leather jacket. After police confirmed King’s identity based on the DNA Doe Project’s findings, Press and Fitzpatrick talked to her mother. They learned that after King went missing, her mom never changed her phone number or address for 37 years, hoping for a visit or call from her daughter. In 2018, she got a call, but it wasn’t the one she wanted.

“We said, ‘At least you got closure,’” Fitzpatrick recalled. “She said, ‘There is no closure. There’s closure for you because you’ll go on to something else, but I’m never gonna go on to something else.’” The DNA Doe Project contributed money to King’s gravestone.

Because of family members like King’s mom, the emotional stakes of solving a Doe case can feel higher to Press and Fitzpatrick, which makes the unsolvable cases all the more frustrating. The DNA Doe Project has been able to close about two in every three cases that come its way. It’s the open cases that sting, that keep the women and their volunteers up at night, restlessly staring into bright screens until their eyelids droop, unwilling to let go.

“The family needs to know what happened... They turn to us because they can’t turn to the policeman and cry.” 

Often their tenacity pays off. It took more than 2,500 hours, for example, for volunteers to solve the case of Joseph Newton Chandler III, whose identification was announced in June.

The process began in 2016, when Cleveland-based US Marshal Peter Elliott read an article that mentioned Fitzpatrick and reached out to her for help. He was trying to identify a man, known as Chandler, who’d stolen the identity of a dead 8-year-old boy in the ’70s. When Chandler killed himself decades later, he left behind $82,000 in a bank account and no known relatives to inherit the money. The case had developed some internet lore around it: rumors of Chandler really being the Zodiac Killer or hijacker D.B. Cooper. Using the Y-DNA method, Fitzpatrick told Elliott the man’s surname was likely Nicholas or a variation of that name, but the lead didn’t go far — not until the DNA Doe Project was formed.

In 2017, Press and Fitzpatrick decided the Chandler case would be their first. They put together a private Facebook group of volunteers — other genetic genealogists they knew and trusted through the online community — and got to work.

Chandler’s DNA was in bad shape, and it would take nearly a year to crack him. But one night, around 1 a.m., Fitzpatrick checked the Facebook group and saw the page was on fire. They’d found him: Robert Ivan Nichols. Fitzpatrick said she called Press, who'd just logged off and gone to bed, and left an urgent voicemail: “‘Margaret, Margaret, I’m leaving a message, get your butt on Facebook, quick, Margaret, call me the minute you get this!’”

They called Elliott the next day with their findings. “It was the first time in US Marshals Service history, since 1789, that we’d ever utilized forensic genealogy, and I had no idea what it was and I’m still not sure I do,” said Elliott, an enthusiastic early adopter of the tool. “It’s phenomenal.”

“It was the first time in US Marshals Service history, since 1789, that we’d ever utilized forensic genealogy.”

But Elliott is also exceptional. Not every law enforcement official understands why this work requires experts. Press said she received an email recently from an agency asking why, instead of paying the DNA Doe Project $1,500, they couldn’t just spend $99 plus shipping to stick their DNA sample in a vial and send it to under an alias. “We have to explain to them — and this is law enforcement — because that’s against the law,” she said. “There’s seven reasons why that’s a bad idea and eight reasons why it wouldn’t work.”

Once, while updating an agency on their progress, Press and Fitzpatrick said they’d identified some third cousins of a Doe. The agency then asked for names, Press said. “We said, ‘You are not gonna go sit on their house. You are not gonna go question these people. They will have no idea what you’re talking about. You don’t know your own third cousins.’ So we have to hold them back.”

Their method is teachable, but not in a workshop or two, which is what some authorities hope for. When Press and Fitzpatrick have tried to teach the technique in a workshop setting at law enforcement agencies, they’ve been met with either glazed eyes or deer-caught-in-headlights looks. In recent months, they’ve heard whispers of departments considering hiring full-time genealogists. Press supports the idea, and Elliott said he predicts it will start happening more, too.

Not long after Press, 71, retired from her computer programming job in 2015 and moved from New England to Northern California to live near her daughter, she realized she finally had time to finish reading Sue Grafton’s iconic alphabet crime series: “A” Is for Alibi, “B” Is for Burglar, etc. Press, who spent the ’90s writing a true crime book and two mystery novels herself, picked up where she left off (around “L”), and when she finished “Q” Is for Quarry — based on the real story of an unknown woman found in Lompoc, California, in 1969 — she was struck by an idea.

By 2017, Press had been a hobbyist genealogist for about a decade, most recently helping adoptees find their birth parents, and she thought she knew a way to identify Sue Grafton’s Jane Doe. She reached out to the Santa Barbara County Sheriff's Office — which didn’t return her calls — and then to Grafton herself, who was interested in hearing more but asked Press to hold off while she finished “Y” Is for Yesterday. It was then that Press contacted Colleen Fitzpatrick, whom she knew only via Facebook as the pioneer of “forensic genealogy,” to brainstorm. This was the beginning of the DNA Doe Project.

But Grafton died before Press could get her hands on Lompoc Jane Doe’s DNA, and before Press and Fitzpatrick could solve their first case. “It broke my heart,” she said. Grafton had spent considerable time and money trying to identify Jane Doe, even exhuming her body so that new facial reconstructions could be made. “I wanted to give her what she wanted most outside her writing,” Press said.

Lompoc Jane Doe still hasn’t been identified, and Press hopes to one day take on the case from the Santa Barbara sheriff. “They probably still have that ‘While You Were Out,’ pink slip on their desk,” said Press, laughing from her own desk, where her computer is framed by layers of colorful sticky notes. Press has soft features and the aura of a well-liked professor with open office hours; she can talk forever and never bore you. And she can’t really blame Santa Barbara authorities for not calling her back. She knew when the DNA Doe Project launched, their idea seemed off-the-wall, unproven, an experiment.

But by July 2018, they'd have three cases officially solved: Buckskin Girl (announced two weeks before the Golden State Killer), Chandler, and “Lyle Stevik,” the man who killed himself in 2001 after checking into a motel with an alias. Even more than Chandler’s case, the Stevik mystery had inspired whole internet communities devoted to identifying him. When the DNA Doe Project set up a “DoeFundMe” to pay for the case, it was fully funded within a day.

But when Stevik was successfully identified, his family was devastated by the news. Though they hadn’t seen him in 17 years, they thought he was alive and well, said the detective-turned-coroner who worked Stevik’s case. The family asked that his real name not be released, for his privacy’s sake and for their own. The case captivated thousands of impassioned people who felt protective of Stevik, and who’d put in work and money to help identify him; what would they say or do to the family that never reported him missing? Many honored the family’s wishes. Others — the “creepy” and “ill-mannered,” Press calls them — are still trying today to uncover his identity through means like the Freedom of Information Act. They feel entitled to it, like they were Stevik’s real family, Press said.

But this aspect of the DNA Doe Project’s work — an attentive audience — hasn’t been all bad. Every one of their cases, before they took them on, had some kind of online following. Chandler/Nichols had his Zodiac message board threads. Buckskin Girl/Marcia King had a Facebook page. This was another big difference between naming suspects and solving Doe cases: The latter had cheerleaders, people who were willing to fund any efforts to identify the unknown, people who were torn up by the tragedy of these unidentified souls.

Everyone fears dying alone, but these Does actually did, and no one in the decades following had claimed them. With no fresh leads, their cases had fallen to the bottom of detectives’ workloads. If no one knew who these Does were, who would look for them? Who would remember them? Strangers on the internet, it turned out.

“You could argue that the greater good is in bringing killers to justice and getting serial murderers off the streets,” Press said. “It’s a much more subtle public good to name the nameless.”

As they solved these cases, DNA Doe Project’s volunteer network blossomed. Press described the group of about 40 as a “think tank.” They work primarily online, and often into the wee hours of the morning, though occasionally they drive to nearby libraries or universities to dig into relevant records. They’re supportive and patient with one another. They aren’t paid by the government agencies they’re assisting; those agencies don’t even know who they are. In their day jobs, they’re private eyes or retired law enforcement or professional adoptee searchers. But this is their downtime, solving genealogy mysteries as they would a Sudoku puzzle.

“We all see how addicting it is, and how you can sit at your computer for hours and can’t give it up,” Press said. “What makes this so addictive is not just getting closer to solving a puzzle, but getting closer to solving a puzzle that people care about, that’s going to matter. It’s going to matter.

By the time a digital file of the frozen baby’s DNA landed in their inbox, ready to upload to GEDmatch, Press and Fitzpatrick had made their decision.

“We told the investigator, ‘We can’t do this case right now, until attitudes have changed and we’ve established ourselves and we’re accepted. This would kill our project. Just kill it,’” Press recalled. She never expected to be making a decision like this. Help name a baby at the cost of sending his mother to prison? Who knew what had led her to that parking lot in January 1988?

“There may be all kinds of reasons,” Press said. “She could be young, she could have been coerced ... She’s still complicit in murder. But it’s not gonna be a poster-child case for getting the community on board with what we’re trying to do.”

Press and Fitzpatrick are experienced genealogists, but they weren’t prepared to consider such weighty ethical life-or-death questions when they started getting into family trees. Now their field is developing so quickly that there aren’t widely agreed-upon ethical guidelines, and certainly not anyone to enforce them. The closest thing to professional standards is a 2015 document drafted by a handful of relatively respected figures in the field. Fitzpatrick doesn’t think it’s nearly enough.

“The best analogy is the people who listen to police scanners and then go when they hear trouble to try and help out and catch a perpetrator who’s running away.”

“Right now genealogy is so self-governed,” she said. “You have these little domains of people who have their own rules, and if they want to violate those rules, it doesn't matter, because they made them.”

Genealogists can be certified by a board and/or accredited by an international commission, but genetic genealogy doesn’t yet have either form of professional validation. This concerns Fitzpatrick and Press, but no one knows how to go about certifying a genetic genealogist: where to start, who to decide it, and what the curriculum might look like.

The result is that without certified professionals, anyone could be considered an amateur — particularly given that most everyone works from home, and that many are retirees doing this work for free, and that genealogy itself usually begins as a hobby, and that the DNA and scientific aspects of genetic genealogy are often self-taught. Fitzpatrick and Press know there are green and unethical searchers in their field. What happens when they convince an agency to send them work?

“I can’t think of anything more inappropriate than to put amateurs to work,” said Nathan Lents, a biology professor at the John Jay College of Criminal Justice.

“The best analogy is the people who listen to police scanners and then go when they hear trouble to try and help out and catch a perpetrator who’s running away,” Lents said. “There’s a reason why it’s best done by professionals. That’s why you have certifications and professional societies — to allow there to be standards of conduct.” The practice of law enforcement commissioning evidence analysis from untrained scientists, he added, is “entirely without precedent.”

But the genealogists working with law enforcement are all outsiders, often referred to as citizen scientists. It’s a small group: There’s the DNA Doe Project and IdentiFinders, the business Fitzpatrick has run since 2011, working on a handful of suspect cases along with other legal matters (heir-finding, fraud detection). Then there’s Parabon NanoLabs, the high-profile company whose public face is media fixture CeCe Moore, spurting out suspect identifications on a seemingly weekly basis. Finally, there’s the team of Barbara Rae-Venter, a more publicity-shy genealogist who identified the Golden State Killer suspect and then kept it a secret for four months.

Recently, the DNA Doe Project volunteers solved their fourth case, a Jane Doe shot to death on a hiking trail near Lake Tahoe in 1982. Press and Fitzpatrick are just waiting for the green light from a sheriff's office to identify her publicly.

“You do need to be really careful. You could destroy someone’s life if you mess up.”

As for the frozen baby, Fitzpatrick has decided to revisit little David Paul’s case independently, with her company IdentiFinders. Press said she’s thrilled; the DNA Doe Project volunteers have plenty of work ahead.

Fitzpatrick believes this will eventually become a “cottage industry, where more genealogists who hopefully have the expertise will be asked by law enforcement to lend a hand,” she said. But the question, as this field grows, is how that expertise will be judged when the industry lacks oversight outside of its occasionally raucous Facebook debates.

Karen Auman, an assistant professor at Brigham Young University — the only college in the country that offers a genealogy degree — said the school will begin offering its first genetic genealogy course next semester, focusing in part on working with government agencies and law enforcement. But she emphasized this is not work for novices: “You do need to be really careful. You could destroy someone’s life if you mess up,” she said.

The women behind the DNA Doe Project acknowledge this too. “Any genetic genealogist worth their salt will do a reasonable job,” Press said. “Our concern is all the ones who want to do it who have no idea what they’re doing.”

“Don’t try this at home,” said Fitzpatrick, sitting in her living room. ●

Topics in this article

Skip to footer