Allison Brocato last saw her sister alive on the afternoon of January 13, 1995. It was a Friday, and Catherine Edwards, Allison’s 31-year-old identical twin, had just gotten off work at Price Elementary School, in Beaumont. On her way home, Edwards stopped to pick up her beagle, whom Brocato had been dogsitting. She lingered a few minutes to chat and to play with Brocato’s infant daughter. “She seemed kind of sad that day,” Brocato would later recall. “I think she had had a fight with an ex-boyfriend the night before.”
Brocato and Edwards considered themselves best friends. After graduating from Lamar University, they both got jobs as public school teachers and moved into a modest town house in west Beaumont, where they lived together until Brocato got married. The sisters looked so alike—a bit shy of five feet tall, slim, with pale skin and shy smiles—that their high school yearbook had mixed up their photos. Later, as teachers, they would occasionally fool their students by pretending to be each other.
The two women spoke again by phone that evening, as they usually did before bed. Edwards had decided to break off all contact with her ex-boyfriend. The sisters both planned to be at the family’s traditional Saturday lunch the next day at their parents’ house, but Edwards never showed up. When her parents drove to her town house to check on her, they found their daughter’s body in the second-floor bathroom, slumped over the tub. She was nude from the waist down, and her wrists were handcuffed behind her back. Her father sounded frantic when he told Brocato what they’d found. “He said, ‘Your sister’s dead, your sister’s dead.’ ”
The brutal murder made front-page news for days in Beaumont, where Edwards was known as a dedicated teacher and a lifelong Presbyterian. She volunteered at St. Elizabeth hospital and served as a mentor for the “I Have a Dream” Program scholarship. Neighbors remembered her walking her dog each evening. “She was loved by everybody,” recalled Steve Thrower, a now retired investigator for the Jefferson County District Attorney’s Office who was assigned to assist Beaumont Police Department detectives with the case. “Great family. Never had any kind of criminal issue. Usually that really shrinks your suspect pool.”
Crime scene investigators saw no signs of forced entry at the town house. Either Edwards had left her door unlocked or she had let her killer in. During the autopsy, a forensic pathologist took a vaginal swab that collected semen, which was also found on the comforter of her bed. The assailant appeared to have raped Edwards and drowned her in the bathtub. Detectives collected DNA samples from dozens of potential suspects, including Edwards’s ex-boyfriend and several of her colleagues at Price Elementary. None matched the perpetrator’s DNA. Nor did the DNA match anyone in the Federal Bureau of Investigation’s recently created Combined DNA Index System, a national database of genetic profiles from convicted criminals, forensic evidence, and missing persons.
As the investigation dragged on month after month without any progress, the case slowly went cold. No arrest was made. For the next quarter century, the semen collected from the Edwards murder sat in a series of evidence-room freezers. Encoded in its DNA was the identity of the killer. But unless he left behind evidence at another crime scene, it seemed unlikely he would ever be caught.
Then, in 2020, a Texas scientist and entrepreneur named David Mittelman approached Beaumont police with an intriguing offer. Two years before, Mittelman had opened a private DNA lab in The Woodlands, an affluent master-planned community north of Houston. He named the company Othram, after the defensive wall of a fortress in J. R. R. Tolkien’s The Lord of the Rings. (“We serve a public-safety function, so there’s a bit of a loose connection there,” Mittelman explained.) The lab specializes in forensic genetic genealogy, a powerful new investigative method that combines whole-genome DNA sequencing with traditional genealogical research based on archival documents such as birth and death records. The technique first came to widespread public attention in 2018, when California detectives used it to identify Joseph James DeAngelo Jr. as the Golden State Killer, a serial murderer and rapist who had eluded police for decades.
When Beaumont detective Aaron Lewallen received Mittelman’s offer to assist with any unsolved crimes, he immediately thought of the Catherine Edwards murder. “This was Beaumont’s most high-profile homicide,” recalled Lewallen, a laconic 26-year veteran who had developed a specialty in cold cases. “With the original detectives ruling out so many of the people who were close to her in her life, it had really become a whodunit.” At Lewallen’s request, local officials agreed to pay Othram about $10,000 to conduct new DNA testing. A few weeks later, a FedEx courier dropped off a Styrofoam box at the lab’s headquarters. Inside, chilled by an ice pack, was a piece of floral-print fabric from Edwards’s comforter and a vaginal swab from the posthumous rape kit.
Unless he was dead or in jail, the man who killed Edwards remained at large. Perhaps he was still in Beaumont. Perhaps he had moved away and started a new life. He had concealed his crime for nearly three decades; surely, he must have thought, the police had given up on the case. There was no way for him to know that in the early 2020s, a small group of detectives and scientists had dedicated themselves to unmasking him.
On a cool gray morning last fall, I drove thirty miles north from Houston to tour Othram’s lab. The company rents space in a four-story building beside a lake in a heavily wooded office park off the aptly named Technology Forest Boulevard. David Mittelman greeted me in the elegantly furnished lobby. Wearing jeans and a rumpled black polo shirt, with disheveled hair and beard stubble, the 43-year-old scientist looked like he had been up all night. He was joined by his 45-year-old wife, Kristen, Othram’s chief development officer. A lean, angular brunette who has moonlighted as a competitive bodybuilder, she was dressed in stylish athleisure wear. The couple met at Baylor College of Medicine, in Houston, where they both earned doctorates—David in molecular biophysics and Kristen in biochemistry.
Together the two have turned Othram into arguably the world’s leading forensic genetic genealogy lab. Over the past six years, the company has been publicly credited with helping to solve nearly 350 cases, including murders, rapes, and unidentified bodies. That number represents only a fraction of the thousands of cases it has actually assisted on, David told me, because some law enforcement officials prefer not to disclose Othram’s role in their investigations. (For comparison, Virginia-based Parabon NanoLabs, another well-known company in the field, says it has assisted in more than 315 cases.) Othram has ongoing relationships with agencies around the world, including the Australian Federal Police, the Royal Canadian Mounted Police, and the Texas Rangers.
Kristen led me through two sets of key card–operated double doors and down a long corridor lined with floor-to-ceiling glass walls. On the left was a series of research labs, where Othram scientists were exploring new ways to extract and analyze DNA. On the right were the forensic labs, where masked technicians in paper gowns and hairnets bent over hooded lab benches, working on crime scene evidence. In one room, a femur, stained dark from many decades underground, sat on a sheet of butcher paper. Because of privacy concerns, the Mittelmans couldn’t reveal anything about the case other than that the bone belonged to a child.
Multiple times a day, packages containing crime scene evidence—blood, bones, hair, nail clippings, teeth—arrive at the lab. An Othram employee photographs the packages and then uploads the images to the company’s online portal, where law enforcement agencies can keep tabs on their evidence. Each step of the process is documented to maintain a chain-of-custody record for use in subsequent legal proceedings. Especially with cold cases, tracking the circuitous route that evidence takes—from a crime scene to a police property room to a forensic-testing lab and back to the property room—is critical. To discredit DNA evidence, defense attorneys will pounce on any potential contamination.
Othram technicians determine whether there is sufficient DNA to build a profile from the forensic material they receive. Unlike the cheek swabs used by medical testing companies, crime scene evidence often contains genetic samples from multiple individuals, and it can include plant or animal DNA. It also deteriorates over time. “No matter how well the investigators try to keep it, it’s organic material,” David explained. “There are things that can develop—bacteria and other kinds of things. So that makes it tougher to read the data.”
The lab ends up rejecting about a third of the evidence it receives. Better to wait until forensic technology improves, the Mittelmans believe, than make a futile bid to obtain a genetic profile. Each test uses up a portion of scarce crime scene DNA. Some forensic labs have destroyed entire samples without obtaining a profile. “In medicine, you would never treat a patient if you had no idea whether it would help or not,” Kristen said. “If you’re running the DNA on assays that don’t work, you’re consuming it, which means you’re consuming someone’s last chance at justice.”
Over time, Othram researchers have developed proprietary methods for obtaining profiles from ever-smaller amounts of DNA. In 2021 the lab established a new milestone by using just 120 picograms of DNA—about fifteen human cells’ worth of genetic material—to help identify the man who raped and murdered a fourteen-year-old Las Vegas girl, Stephanie Isaacson.
In the Edwards case, Othram technicians determined that the semen found on the vaginal swab had a high likelihood of yielding a strong DNA profile. They used a technique called differential extraction to distinguish the suspect’s sperm from Edwards’s skin cells and other foreign material. Then they ran the sample through an Illumina NovaSeq 6000, a million-dollar whole-genome sequencer about the size of an office copier. Othram’s custom software combed through the data for genetic markers—unique DNA sequences that could be used to identify the suspect. To build a useful profile, Othram investigators need to find hundreds of thousands of such markers. In the Edwards case, they found more than half a million.
The genetic profile they developed was then uploaded, in the form of a digital text file, to a website called GEDmatch. Founded in 2010, the site maintains a database of more than 1.5 million profiles voluntarily submitted by users around the world, many of them hoping to find distant relatives. About 30 percent of those users have consented to law enforcement using their personal information to identify violent criminals. The website says it has helped solve more than four hundred cases.
The closest genetic match to the suspect in the database was a woman living in Louisiana. Based on their quantity of DNA in common, they appeared to be second cousins, which meant they shared a pair of great-grandparents. The woman might never have met the suspect or even known of his existence, but she had just become an unwitting genetic informer for the Beaumont Police Department. The information she provided was about to break the Catherine Edwards case wide open.
David Mittelman’s path to Othram began in 1997, when he landed an internship at the University of Texas Southwestern Medical Center, in Dallas, working on the Human Genome Project—the thirteen-year, $3 billion effort to map our entire genetic composition. At the time, he was a precocious student at Pearce High School, in nearby Richardson. “It just seemed like an exciting opportunity to learn more about what all these pieces of genetic information meant,” Mittelman said. “Intuitively, I could sense that this was going to be a real driver in changing how society works.”
He continued working on the project while attending college at UT Dallas. His job was to build and program robots that would automate various lab processes. “Genomics in the nineties was very labor-intensive,” he explained. “Back then, it was a lot of human stuff and a little bit of computer stuff. That has kind of flipped now.” After earning a bachelor’s degree in neuroscience from UT Dallas and his doctorate, David took a professorship at Virginia Tech, where he won a National Institutes of Health grant to research the use of genetic data for medical diagnosis and treatment.
One day in the early 2010s, a representative from the FBI visited Virginia Tech to speak about the role of DNA in criminal investigations. Humans share roughly 99.9 percent of our genetic material, but all of us have DNA sequences—typically along portions of the genome whose functions aren’t yet well understood—that feature enough mutations to distinguish us as individuals. By the mid-1980s, scientists had identified the location of thirteen such sequences, also known as markers, which became known as a DNA fingerprint. (The number has since expanded to twenty markers.) Because the odds of any two unrelated individuals sharing the same DNA fingerprint are infinitesimal, this test has become the standard in international law enforcement. But the method works only if police have a suspect from whom to collect a DNA sample. If the culprit remains at large, as in the Catherine Edwards murder, the method is all but useless.
During the discussion, David realized that the field of forensics was woefully behind the times. Rather than relying on twenty markers, scientists could now use whole-genome sequencing to obtain hundreds of thousands. Sequencing an entire genome would give police detailed information about a suspect’s ethnic background, eye color, sex, and skin pigmentation. It would be like going from monochrome to Technicolor. “The FBI guy asked me how much it would cost, and I said thirty thousand to forty thousand dollars,” David recalled. “He laughed at me and left. At that time, the economics didn’t make any sense.”
David set aside the idea and continued working on the medical side of genetics. In 2012 he cofounded Arpeggi, a tech start-up that created software to help physicians sift through the avalanche of data produced by whole-genome sequencing. Within a year the company was acquired by Gene by Gene, the Houston-based parent company of the popular genetic testing website FamilyTreeDNA. David resigned from Virginia Tech and moved back to Texas to become Gene by Gene’s chief scientific officer. This gave him an introduction to the world of direct-to-consumer DNA testing, which had exploded in popularity. Although he soon left to work at a series of other biomedical companies, his experience at Gene by Gene planted the seed for Othram.
“The costs were coming down on sequencing,” he said. “I had learned a lot at FamilyTreeDNA about how people interconnect genetically. It was, like, maybe the economics are where they need to be. The technology has largely been solved for medical diagnostics. So how can we take this powerful technology to do something good in the world? And that’s what shifted my mindset back to the problem of forensics.” David soon discovered that there were hundreds of thousands of unsolved murders and tens of thousands of unidentified bodies in the United States alone. Crime labs nationwide were struggling to keep up with new investigations, let alone decades-old cold cases. The more David learned, the more outraged he became. “People do not understand the magnitude of this problem,” he told me.
In 2018, with $4 million in seed money led by a San Francisco–based investment fund, David set up shop in a one-story office building in The Woodlands, a five-minute drive from his house. Kristen initially declined to take a job at the company because it seemed like such a long shot. “I told him that he had lost his mind,” she said. “Who is going to give you evidence? We’re medical people. We have no policing background whatsoever. And you’re going to build a forensic lab?”
Genetic genealogy is such a young field that few regulations or accreditation standards exist. David knew that he would have to win the trust of law enforcement agencies unfamiliar with whole-genome sequencing. He would have to convince them that such testing was worth the initial price tag of $10,000 per DNA sample. Perhaps most important, he would need to prove that his work would stand up in a courtroom.
Brandon Bess, a bluff, plainspoken Texas Ranger who lives in the small town of Anahuac, midway between Beaumont and Houston, visited the lab shortly after it opened. “It was David and, like, three other people working there at the time, and it was kind of a dump,” he recalled. “My first impression of David was that he looked like a mad scientist. He had on a T-shirt that was too small and looked like he hadn’t slept in about three days. He had hair going all over the place, a ripped-up pair of jeans, and a pair of tennis shoes. But he was very focused. And he could talk.” (David told me Bess’s impression was accurate—early on at Othram, he would often work for days with little sleep.)
Whole-genome sequencers such as the NovaSeq 6000 were built to analyze fresh DNA obtained from cheek swabs. Othram had to design lab processes capable of extracting data from damaged and degraded genetic material, then create software to analyze it. David hired engineers and molecular biologists with experience pulling DNA from inhospitable media, such as formaldehyde and paraffin, which are known to scramble genetic information. “We developed a number of tools, both in the laboratory and on the computer side, that allowed us to get reproducible and predictable success from forensic samples,” he said.
In 2019 Othram built a DNA profile that enabled investigators to identify a skeleton discovered in an Idaho cave in 1979 as the remains of Henry Loveless, a bootlegger and counterfeiter who was murdered in 1916. Then, in 2020, the lab received a flood of publicity after helping to solve the 1974 murder of Carla Walker, a seventeen-year-old Fort Worth high school student.
The company soon began announcing new successes on an almost weekly basis. It helped identify the perpetrators of the 1974 abduction and murder of a five-year-old girl from Montana; the 1977 rape and murder of a 77-year-old North Carolina woman; and the 1984 rape and murder of a nine-year-old girl in Canada. There were also many John and Jane Does it helped identify. A victim of a traffic accident on a Pennsylvania road in 1987. A corpse fished out of a lake in Washington State in 1994. A woman who mysteriously drowned in a Pecos hotel swimming pool in 1966.
To publicize Othram’s work, Mittelman capitalized on America’s fascination with true crime. He became a regular speaker at CrimeCon, the popular annual true-crime conference, and he consulted on a 2021 episode of Law and Order: Special Victims Unit that featured a forensic genealogy lab inspired by Othram. The company hosts a private genetic database where users can upload their DNA data in hopes of helping solve a crime, and it has launched dozens of crowdfunding campaigns to finance work on cold cases around the country. On social media, Othram advertises its successful investigations and solicits donations.
Among the company’s earliest large donors was Carla Davis, a self-taught genealogist from Mississippi who came across one of Othram’s fund-raising campaigns on LinkedIn in late 2020. She gave tens of thousands of dollars that helped fund the identification of several sets of remains found in Mississippi between 1977 and 2020. “Law enforcement agencies there are so underfunded,” Davis recently told me over Zoom from her home in Dubai, where she and her husband have a real estate investment company. “I was just trying to help give my home state the technology to solve their cases.”
After Davis continued to lend Othram both financial support and genealogical expertise, the company hired her two years ago to lead its thirteen-employee genealogy team. Working remotely, Davis and her team have assisted in more than four hundred cases. “It’s really exciting to be part of this moment,” she said. “We’re going to see so many more companies using this technology and more cases being solved. I think we’re not going to have cold cases in the future.”
According to the business-research platform Crunchbase, over the past six years Othram has raised nearly $36 million in venture capital. Among its biggest investors is Gigafund, the Austin-based firm best known for its stake in several of Elon Musk’s companies, including the Boring Company, Neuralink, and SpaceX. “Our investors are interested in the decade-long transformational shift in the way forensic genetic testing is done,” David said. “They understand it’s a long game.” Like many tech start-ups, Othram has spent its early years burning through money in pursuit of its goals. In 2022 it moved to another office building in The Woodlands—with quadruple the lab space and additional room for a staff that now numbers more than sixty. It currently charges law enforcement agencies only enough to cover the incremental costs of an investigation, though David said that won’t remain the case as economies of scale bring those costs down. He also plans to reach profitability by licensing the company’s software and processes to other forensic labs.
Othram has attracted criticism for what some consider its sharp-
elbowed business practices. In January, the company announced an exclusive partnership with FamilyTreeDNA, one of only two major databases that grant law enforcement agencies access to their profiles. (The other is GEDmatch, which is owned by Netherlands-based biotechnology company Qiagen.) Labs hoping to use FamilyTreeDNA to identify a suspect now have to use Othram software. In the wake of the announcement, the nonprofits DNA Doe Project and Intermountain Forensics temporarily stopped using FamilyTreeDNA. “They were seeking additional clarification, so I personally got on the phone with them to talk them through the partnership,” Mittelman said. Both organizations have since resumed working with FamilyTreeDNA.
Meanwhile, the field of genetic genealogy has come under fire from civil liberties groups concerned about the privacy of users who upload their DNA information to websites such as GEDmatch, and from bioethicists who worry about the dearth of federal regulation. DNA is valuable—to governments, to bioscience companies, and to the police—and genetic databases, like anything else that lives online, can be hacked. Mittelman says he shares some of these concerns but points out that users submit their DNA to these sites voluntarily and can opt out of law enforcement searches.
Forensic genetic genealogy relies on a certain degree of ethical ambiguity. After a lab such as Othram identifies a potential culprit, detectives must then collect the suspect’s DNA—often surreptitiously—in order to match it to DNA found at the crime scene. Usually this is done by taking items from a suspect’s garbage, as in the Golden State Killer case. According to long-standing legal precedent, police typically do not need a warrant to obtain evidence from garbage that has been left on the curb. So far, courts have ruled that the same holds true for the DNA in that trash.
Some experts, including Teneille Brown, a law professor at the University of Utah who has written about forensic genetic genealogy, argue that courts should declare the furtive seizing of someone’s DNA to be a violation of the Fourth Amendment’s prohibition against “unreasonable searches and seizures.” Obtaining someone’s entire genetic profile, after all, constitutes a significantly greater invasion of privacy than merely rummaging through their food scraps. “DNA is not garbage, so we shouldn’t be treating it like that,” Brown told me. The American Civil Liberties Union has filed several amicus briefs opposing warrantless DNA collection.
Mittelman professes not to concern himself with such constitutional questions, deferring to the legal experts. For him and most law enforcement agencies, what matters is the ability to identify a victim or a suspect. At that Othram has proved remarkably successful. During one week in late May, the lab announced its role in solving six cases, including one murder and five unidentified bodies. And it is increasingly called upon to assist in active investigations. Earlier this year, Othram helped identify Victor Antonio Martinez-Hernandez, a fugitive from El Salvador, as a suspect in the 2023 murder of Maryland mother Rachel Morin. Martinez-Hernandez was arrested in June and charged with murder and rape.
With genetic-testing technology improving rapidly, the biggest remaining hurdle to solving cold cases may be money. “These cases move at the speed of funding,” David told me. In the United States, the largest single source of law enforcement funding is the federal government. This is where Kristen Mittelman comes in. After earning her doctorate, she worked as an intellectual-property specialist at Houston law firm Baker Botts before moving with David to Virginia Tech, where she became the university’s director of grants and contracts. In 2021 Kristen joined Othram, where she spends much of her time in Washington, D.C., lobbying Congress for money. “He was struggling with funding,” she told me. “And I knew I was good at getting federal funding.”
Thanks in part to Kristen’s efforts, Republican congressman Kelly Armstrong, of North Dakota, introduced the Carla Walker Act in 2022. Named after the Fort Worth teenager whose murder Othram helped solve, the bill would provide $20 million in grants to law enforcement agencies for forensic genetic genealogy investigations. The bill has yet to make it out of committee. Last August, U.S. senator John Cornyn traveled to The Woodlands to meet the Mittelmans and tour Othram. During a press conference at the lab, he promised to introduce a version of the Carla Walker Act in the Senate. “This sort of technology is critical to solving crime and protecting public safety,” he declared.
Kristen is working with Cornyn’s office to draft the proposed legislation, which she says will include reporting requirements to assess the work of forensic genetic genealogy labs. The Mittelmans worry about the damage an unscrupulous lab could do to the reputation of their nascent industry. “There are so many companies out there selling the quickest way to make a profile with fewer markers,” Kristen said. “Because there are currently no metrics of success in forensics, people are taking shortcuts. This needs to be done predictably. You need to start collecting metrics, find technologies that work, and fund the implementation of those technologies.”
After identifying the likely second cousin of Catherine Edwards’s killer, Othram passed its findings to Beaumont detective Aaron Lewallen, who was now working with Brandon Bess, the Texas Ranger. Othram has an in-house genealogy team, but some law enforcement agencies use their own genealogist. In this case, Lewallen happened to know one willing to work for free: his wife.
Tina Lewallen, a detective in the auto crimes division of Beaumont Police Department, first got interested in genealogy in the nineties, to explore her family history. When direct-to-consumer DNA testing became available, she submitted a cheek swab to AncestryDNA. The results ruled out the French heritage her family had always claimed. “That fascinated me, because DNA doesn’t lie,” Tina recently told me in her windowless Beaumont office, which was decorated with mug shots of suspected car thieves. A pair of rose-pink handcuffs, a gift from her husband, were clipped to her belt.
Othram provided Tina with a list of the suspect’s closest relatives on GEDmatch, along with how much DNA each shared and their likely familial relationship. The listed email address for one of the relatives, Paul Thomas LaPoint, led to his daughter-in-law, a professional genealogist named Shera LaPoint who lives in the small town of Bunkie, Louisiana, about eighty miles northwest of Baton Rouge. Shera had submitted Paul Thomas’s DNA to GEDmatch years before. Stunned to learn that her father-in-law was related to a suspected killer, she volunteered to help the Beaumont detectives with the case. After vetting her, the Lewallens brought Shera onto the team, forwarding the list they received from Othram.
“It was a bunch of Cajun names, many of them people that I knew, because a lot of them are also into genealogy,” Shera told me. Forensic genetic genealogy investigations start by identifying the most recent common ancestor between the suspect and their closest genetic match. In the Edwards case, the search was complicated by the suspect’s Cajun ancestry. Cajuns descend from a small colony of French Canadians who were expelled by the British in the mid-1700s and found refuge in Louisiana. Over the centuries, some members of the tight-knit Catholic community engaged in intermarriage with close relatives, such as cousins. As a result, performing genetic genealogy in Cajun families can be notoriously complex. Shera and her father-in-law, for instance, share thirteen ancestors. “In a perfect world, when you look at DNA matches for a person who has tested on AncestryDNA or FamilyTreeDNA, you should be able to separate their four grandparents’ lines,” Shera said. “But when you’re looking at Cajun DNA, it’s very difficult, because your maternal grandmother may be related to your paternal grandfather. So it makes it very difficult to find the common ancestor you’re looking for.”
Using Ancestry.com, which bills itself as the world’s largest genealogy site, Shera and Tina built a family tree for the presumed second cousin, going back in time until they identified her eight great-grandparents. Then they started working back down, following branches of the tree in search of a descendant who lived in Beaumont when Catherine Edwards was killed. But that effort led to a dead end. Because the family was Cajun, they realized, the presumed second cousin might actually be the suspect’s third cousin. Shera and Tina were forced to go back another generation, to the woman’s great-great-grandparents. They ended up with a family tree of more than 7,400 names.
To narrow the search, they asked Aaron Lewallen and Bess to request DNA samples from living members of these families. The detectives spent days driving around Texas and Louisiana, collecting dozens of cheek swabs. “I thought it was going to be difficult to talk people out of their DNA,” Aaron recalled. “But you’d be amazed how many people are out there interested in helping out. There’s a lot of true-crime buffs.” Each swab was sent to Othram, which sequenced the DNA and uploaded it to GEDmatch. The results let Shera and Tina rule out entire family lines.
About three months into the investigation, Shera’s research led to a husband and wife who’d lived in Beaumont in the sixties. Birth records indicated that the couple had two sons who would have been about the same age as Edwards. Shera texted Tina and Aaron, who ran the names of the couple’s sons through a criminal background search. One of the brothers came up clean, but the other had a record. In 1981, Clayton Bernard Foreman had pleaded guilty to aggravated assault in Beaumont. “I was like, holy shit,” Aaron recalled. “It’s him.”
According to the case file, Foreman, then a 21-year-old Nabisco salesman, had been driving through Beaumont when he saw a young woman who’d had car trouble. Foreman stopped to offer her a ride, claiming to be a cop. He drove the woman to a secluded area, threatened to cut her throat with a knife, tied her hands behind her back with a belt, and raped her. About two weeks later, the traumatized woman went to the police. Foreman readily confessed, explaining that he had “been out drinking and just got carried away.” In exchange for pleading guilty to aggravated assault, he received three years of probation. Aaron soon learned that Foreman and Edwards were three years apart at Forest Park High School. Edwards and her twin sister, Allison, had even been bridesmaids at Foreman’s 1982 wedding.
An online search revealed that Foreman, now 61, was living with his fiancée in a suburb of Columbus, Ohio, where he worked as an Uber driver. An officer from the local police department was dispatched to collect bags of trash from outside Foreman’s house. Several of the discarded items, including dental floss and plastic tableware, were sent to the Texas Department of Public Safety crime lab in Houston for testing. Foreman’s DNA was a clear match to that of the semen from Edwards’s vaginal swab.
On April 29, 2021, Aaron and Bess flew to Ohio, where they met Foreman face-to-face at the local police station. A video recording of the interrogation shows an obese man with a receding hairline, thick glasses, and a pronounced East Texas accent. At first the detectives act as though they are merely seeking information about Edwards’s murder. Foreman says he vaguely remembers her being a bridesmaid at his wedding but denies having any other contact with her or even knowing that she is dead. Only after being told that his semen had been found in her body does Foreman stop talking and ask to see a lawyer.
Aaron and Bess allowed Foreman to leave the interrogation room unimpeded. On his way out of the station, he was stopped and handcuffed by police officers. Foreman may have recognized the handcuffs. They had been sitting in an evidence room for nearly three decades, until Aaron and Bess received special approval to take them to Ohio. They were the same pair that had been found on Edwards in 1995.
After a three-year delay caused in part by the COVID-19 pandemic, the Clayton Foreman trial began on March 11 in downtown Beaumont. The visitors gallery was packed. Foreman sat calmly beside his attorney at the defense table, wearing an inscrutable expression. He had pleaded not guilty; if convicted, he would face a mandatory life sentence. Prosecutors decided not to seek the death penalty.
One of the first witnesses was Allison Brocato, now sixty years old with shoulder-length hair and matronly glasses. She spoke about her closeness with her twin sister, noting that she and her husband named their second daughter Catherine. After the murder, she said, the family was never the same. “I think my parents died a little bit, too, when she did,” Brocato told the jury, through tears. Neither lived to see the arrest of their daughter’s killer.
The prosecution also called Foreman’s former wife, Dianna Coe, who testified that she had learned about Foreman’s rape charge just three weeks before their wedding. When she confronted him, he claimed that the arrest was “a big misunderstanding, and that the charges had been dropped.” Coe, who was nineteen and had been dating Foreman for three years, decided to go ahead with the wedding. In 1984 she and Foreman had a son, who later attended Price Elementary while Edwards was teaching there.
One day Foreman confessed to his wife that in high school he felt protective of Edwards and Brocato. “He thought they were so cute because they were twins,” Coe testified. “I didn’t think anything of it at the time.” Sometime around 1987, Coe discovered a briefcase in the trunk of Foreman’s car. Inside it was a gun, a pair of handcuffs, and pornographic material. The couple divorced in 1993. Years later, when Coe learned that Edwards had been murdered, she called her ex-husband to tell him the news. “He had no feeling whatsoever,” she recalled. “It dumbfounded me.”
David Mittelman took the stand on the third day of the trial. Wearing a white shirt and blue blazer, with a relatively clean-shaven face and hair that looked recently barbered, he patiently walked the jury through Othram’s role in the Edwards case. Although the lab’s work has been used to convict dozens of murderers, this was Mittelman’s first time to testify before a jury. He explained that Othram had found more than half a million genetic markers in the crime scene DNA, “far in excess of what is necessary to produce a workable profile.”
After seven days of testimony, the case went to the jury. It had taken nearly three decades to identify a suspect in the murder of Catherine Edwards, but it took the jury less than an hour to find Foreman guilty. He received an automatic life sentence and will be eligible for parole in thirty years, when he is 93. After the trial, one of the jurors gave an interview to a Beaumont TV station. He said it was the DNA evidence that convinced the jury of Foreman’s guilt: “You can’t deny that.”
In the wake of the verdict, Mittelman expressed pride in having helped to resolve the three-decade-long investigation. “Unsolved cases take an immense toll on families,” he told me. “Knowing that our technology has played even the smallest role in bringing both answers and then justice is profoundly moving.”
The future is already here,” novelist William Gibson once observed. “It’s just not very evenly distributed.” David and Kristen Mittelman believe that forensic genetic genealogy will one day become as commonplace as fingerprint analysis. For now, this investigative tool still suffers technical limitations, including the relatively homogeneous geographical origin of the DNA profiles on GEDmatch and FamilyTreeDNA, most of which come from Australia, Europe, and North America. The majority of forensic genetic genealogy teams are located in the United States.
Considerable need exists for this technology. As crime spiked around the country during the COVID-19 pandemic, law enforcement agencies were tasked with investigating a tide of new offenses—more than one million violent crimes a year in the U.S.—with a steady or shrinking number of officers. Every crime that isn’t solved adds to the growing number of cold cases. In addition to hundreds of thousands of unsolved murders in the U.S., there are countless rape kits sitting in evidence rooms nationwide. For a variety of reasons, including insufficient funding, many have not received any kind of DNA testing.
Even when a kit does get tested, nearly always for the standard twenty genetic markers in a DNA fingerprint, it often doesn’t match any of the profiles in the FBI’s Combined DNA Index System. In those cases, police must either wait for the rapist to strike again—leaving DNA at another crime scene—or pay for forensic genetic genealogy. Although costs are coming down, the method still runs about $10,000 per case, in addition to the expense of hiring a skilled genealogist. America spends more than $100 billion every year on law enforcement, but little of that is earmarked for forensic genetic genealogy. “The amount of money it costs to investigate a case using traditional methods is absurd,” David said. “And the vast majority of it goes to salaries.”
That’s why Othram sees government support as essential. Only Congress has the resources to fund genetic genealogy work at scale, and only Congress has the power to encourage local police departments to make use of it—perhaps by threatening to withhold federal grants. (In the eighties, the federal government compelled states to raise the drinking age to 21 by indicating it would hold highway funding hostage.) But to be eligible for such funding, genetic genealogy will likely need to emerge from its Wild West era and embrace regulation. David and Kristen told me that Othram would welcome such a change. “For a technology to be successful, it has to be predictable,” Kristen said. “That’s how medical testing works. That’s how almost everything works that is funded by the government.”
Like many start-up founders, David balances his frustration at present circumstances with a supreme confidence about what’s yet to come. He predicted that the technology pioneered at his modest lab in The Woodlands will one day become standard in police departments throughout the world. “There was a sentiment, especially in the early years, that cases like Catherine Edwards’s were remarkable one-off successes—things that are extraordinary,” he told me. “We’re trying to go from the extraordinary to the ordinary.”
This article originally appeared in the August 2024 issue of Texas Monthly with the headline “Decoding a Killer’s DNA.” Subscribe today.
