COLUMBIA, S.C. — He attacked in the dark. Striking in the early morning hours, the rapist targeted women and girls who he knew were alone, brazenly breaking into apartments and townhomes around Spartanburg to terrify, assault and abuse.

From 1995 to 2003, he raped or attempted to rape 13 girls and women, with the victims ranging in age from 14 to 51. But after 2003, the attacks appeared to stop, as if the attacker slipped away into the night.

Police were stumped. A composite sketch and DNA samples led to no arrests. Months, years and nearly two decades passed as the investigation was no closer to catching the rapist.

But in 2019, Spartanburg police went looking for help in an unexpected place, the county library.

Charity Rouse, the director of local history at the Spartanburg Public Library, was brought onto the case. A professional librarian, Rouse’s main expertise is as a genealogist.

And in just six months, they had their man: Gregory Frye.

They did this through an emerging field known as genetic genealogy. Rouse and investigators used DNA left behind by the rapist to identity distant relatives and then — pairing time tested genealogical research methods like combing newspaper archives and vital records, with targeted DNA tests of possible relatives — they built out a complex family tree that led them to Frye.

Frye matched a composite sketch, lived in the same apartment complex as the first and fourth victims and, most importantly, his DNA matched semen taken from the scene.

It’s proof of how a “proactive” approach to DNA can break open unsolved crimes, said Mark Hillers, captain of the Spartanburg Police Department’s Criminal Investigations Division. Were it not for genetic genealogy, Frye, a 57-year-old one-handed mechanic who had only been charged with minor crimes like passing a bad check, marijuana and driving violations, likely would never have been caught.

In June of this year, Frye, who pleaded not guilty, was tried and convicted of burglary in the first degree, criminal sexual conduct in the first degree, and kidnapping. While prosecutors had DNA that linked him to six assaults, the decision was made to streamline the process and try him on a single case. It was the first time that a case based on genetic genealogy ever went to trial and won a conviction in South Carolina, according to prosecutors.

Genetic genealogy has emerged as an important new tool in law enforcement, revolutionizing cold cases and giving investigators a new, proactive way to use DNA evidence that might not otherwise have matches.

The technique gained a degree of international attention, and notoriety, when investigators in California used genetic genealogy to track down Joseph James DeAngelo, better known as the Golden State Killer, a serial killer and serial rapist responsible for at least 13 murders and 51 rapes in California from 1974 to 1986.

“Can you imagine standing out there for years wondering if that person is still out there and now knowing they’ve been brought to justice?” said 7th Circuit Solicitor Barry Barnette, who prosecuted the case against Frye along with Assistant Solicitor Lindsey Overby. “I think it would bring a huge weight off of any of their lives. They didn’t know where he was. Now they know exactly where he is.

The attack

Sharon Emory had been out with friends the night of April 3, 1997. She’d had a few drinks, and when she made it home to the Birds Nest Apartments in Spartanburg, she called her ex-husband in Florida and spoke to him for an hour before falling asleep.

While The State does not typically identify victims of sexual assault, Emory testified at trial and gave permission to use a variant of her name.

Just after 4 a.m. she awoke and saw Frye kneeling next to her bed

The light was on and she could see his face clearly.

“I just froze. I just looked at him for a while,” Emory said. “He panicked and put the blankets over my face.

“I felt like I was stronger than him and he just kept attacking me and then started threatening me, wanted me to keep still and threatening me.”

In the struggle, Emory ended up on her back with Frye on top of her.

“He put a pillow over my face and I begged him to let me breathe. He said, ‘If you don’t stop fooling around I’m going to end your life.’ And that’s when everything happened.”

After Frye left, Emory panicked. She tried to call her ex-husband back; when she got his roommate on the phone he told her to call 911. She grabbed a golf club for protection while she stood there, scared and shaking.

Immediately following the attack, officers collected DNA evidence, and Emory worked with a sketch artist to help develop an image of her attacker. But then nothing happened. There were no breaks in the case and she later learned that he’d previously attacked another woman at the same apartment building.

Over the years she underwent therapy and counseling, but the fear and sense of stigma were hard to escape.

“I was just looking over my shoulder, being paranoid and just not myself. You feel violated,” Sharon said.

Twice, a detective came to her work site to ask her to look at a photo lineup, but it just left her feeling embarrassed and exposed. “It was right in the middle of work, I said, ‘right, now the whole work knows about it.’”

As time passed, Emory said that she tried to move on with her life. She got married again but her second husband died. She moved away from Spartanburg and adopted three large, aggressive German Shepherds to help her feel safe.

But the fear, Emory said, always lingered.

The genetic genealogist

Charity Rouse is not who comes to mind when you think of a detective. A genealogist and librarian by training, Rouse says she doesn’t even like true crime shows. She finds them “too graphic.” Her days are spent in the Spartanburg Public Library’s Kennedy Room, a repository for the region’s local history, helping visitors chase down far flung ancestors — not serial criminals.

Her interest in genealogy was sparked by mysteries in her own family. While her mother’s side of the family is huge — Rouse said her maternal grandmother was one of over 104 first cousins while her grandfather was one of at least 73 — her father was orphaned at a young age, raised by a maternal aunt with very little contact with the rest of his family.

And she’s no stranger to how the tragedy of unsolved crimes can linger. In 1996, Rouse’s childhood friend and neighbor, Alicia Showalter Reynolds, was abducted and killed while driving on a state highway near Charlottesville, Virginia. Her body was found two months later, but the killer has never been caught.

“It’s been more than 25 years now and I still get a little choked up,” Rouse said, recounting how the two attended the the same high school together in Virginia and played oboe together in the band. “It’s just one of those situations where you never think it’s going to happen to you, but you always hope that if something tragic happens, there are answers.”

When the Spartanburg Police Department first approached Rouse in October 2018, all it wanted was to see whether she could help explain to investigators and analysts how the genealogy process would work.

Over the years, investigators had tried and failed to match any one of the six DNA samples taken from the crime scenes with a suspect.

They were also puzzled by the sudden stop of the crimes and the fact that no other crimes were connected to the rapist by DNA. The assumption, Hilllers said, was that he had died.

“Because it typically doesn’t stop and then restart,” Hillers said, adding he’s not sure Frye would have ever restarted committing rapes. “This guy would never have been arrested for anything that would have required him to provide DNA,” Hillers said.

Working with the samples that they had, the police had contracted with a private genetic genealogy company to help them get the analysis started, but contracting with them to do the full investigation would have cost “an arm and a leg,” Rouse said.

At first Rouse was just teaching law enforcement basic principles of genealogy, like she would any class at the library or a local historical society. But over time, her role grew from a sounding board who could explain genetic concepts into a consultant working alongside analysts and investigators to help build out the family tree that would lead to the suspect.

But other than some basic information of the rapist, including his race and approximate age, investigators kept most of the details about the case from Rouse.

“Law enforcement was very careful not to tell me much about the case itself,” Rouse said. “It protects me and I don’t need to know that information.”

The hunt

Working with Caitlyn Horn, an analyst at the Spartanburg Police Department, Rouse’s mission was to uncover the family tree that might link a person with the unknown rapist.

Normally with genealogy, researchers start from a known person and build backwards into the past, discovering unknown relatives and ancestors. With genetic genealogy the process is inverted. By starting with known but distant relatives, genetic genealogists build out a tree to lead to the identification of one person.

“Because you’re doing all of this blind, this is like doing adoption research,” Rouse said. “You have the DNA test (of the suspect), but you know nothing about the family around them. You’re building the tree backward.”

From the start, investigators caught a lucky break. Once the suspect’s DNA had been uploaded to a commercially available platform, populated by at-home DNA tests submitted by people all around the country, they found someone who appeared to be a second cousin, Rouse said. That meant they needed to work back to identify great grandparents and then start identifying family lines that split off from there, according to Rouse.

But Rouse and Horn still had to play a six-month long game of “hot and cold” to identify the rapist. They built out a family tree, using news clippings, obituaries and public announcements to identify the names of relatives and discover new branches and “trunk lines.” As well as identifying names, Rouse and Horn would look to see where families were located and whether people could have been in the area when the crimes were committed.

In order to make sure that they were heading in the right way, law enforcement investigators would be dispatched to request DNA samples from possible relatives.

The more closely two people are related, the more DNA they have in common. Commercial DNA testing platforms, like the one that Rouse and Horn used, measure this closeness in a unit called centiMorgans.

The higher the number of shared centiMorgans, the closer two people are likely to be related, with specific degrees of relationships being associated with certain centiMorgan ranges.

While the numbers range and there are always outliers, Rouse said, shared centiMorgans increase as the relationship gets closer. First cousins share around 900 centiMorgans, siblings share approximately 2,400 and parents and children share around 3,500 centiMorgans.

Sometimes, after investigators obtained a DNA test, the number of shared centiMorgans showed that they were moving further away from the suspect. But like all good science, Horn said that negative results were not a failure but actually helped narrow the search.

“It’s not really a setback, because if there’s not a high familial relationship, then it excludes that branch,” Horn said.

DNA was ultimately collected from six people before investigators identified someone they believed to be the rapist’s sibling.

When investigators tested the suspected sibling’s DNA against the sample from the rapist, the centiMorgan match was overwhelming.

“The number was so high, it was only possible for one familial type of relationship,” Horn remembered.

After Frye was arrested, the South Carolina Law Enforcement Division compared a court-ordered swab taken from his cheek with the original sample. It was a match.

According to the SLED report, the probability of a randomly selected individual having the same DNA profile was 1 in 720 quadrillion.

Pins and needles

On April 3, 2019, 24 years after his first known attack, Frye was arrested by officers with the Spartanburg County Sheriff’s Office.

It was a watershed moment for police in South Carolina, which law enforcement officials have said they hope shows the potential of genetic genealogy to solve cold cases and the most serious crimes. But concerns over privacy implications linger.

States, including Maryland, Montana, Utah and Texas, have passed laws to place guardrails on how law enforcement agencies can use commercially-available or open source DNA databases for their investigations.

Genetic genealogy by law enforcement has been compared to warrantless searches, and some genealogy platforms have started cracking down, making less information publicly available to researchers and law enforcement.

“The legal questions raised by genealogy searches are measurably simpler than the ethical concerns,” read one paper published in the National Institute of Health.

“Some people have gotten a little leery about having their DNA available to match for law enforcement testing,” Rouse said. Sensitive to concerns over privacy, many companies have restricted public and law enforcement access to people’s family charts and now require users to “opt in” to law enforcement access instead of an “opt out” system. As a result, Rouse said that the data pool for investigative genetic genealogy has shrunk.

“With new technology, you’re always going to have pushback, just like there’s pushback with license plate reading technology, facial recognition technology,” said Hillers, with the Spartanburg police. “The assumption is always that Big Brother or the government is using it in an evil type of way and we have to combat that.”

Genetic genealogy is expensive and involves huge amounts of time and labor. It isn’t used in run of the mill cases, Hillers said.

“Most of our (rape) cases are known acquaintances, where there’s a question on consent,” Hillers said. “It’s very rare that we get cases where you have an unknown individual enter a stranger’s location, commit an act like rape. Honestly. I know that’s everyone’s biggest fear, but statistically, that is not what happens.”

Barnette said genetic genealogy is “a great tool for cold cases.” But the solicitor recognized that searching for criminals in a family tree can cause real pain for family members, far greater than the occupational hazard of unexpected half siblings or other familial secrets that Rouse says she might run across during normal genealogical research.

“That is something that we try to keep in mind with these kinds of cases,” Barnette said.

For Sharon Emory, the fact that genetics was able to solve her case after all these years was “amazing.”

While Frye’s DNA matched on all six cases where investigators had DNA, by the time of the trial, two of the victims whose cases had DNA had died. Barnette said that the decision was made to take just Emory’s case to court because it had the most straightforward DNA evidence.

Emory said that she was on “pins and needles” leading up to her testimony and was overjoyed when the guilty verdict came in. “I couldn’t believe it, he’s gonna go in. I was happy it was all done.”

Frye was convicted by the jury in two hours, following a three-day trial. He received a life sentence for burglary and 30 year sentences for kidnapping and first degree criminal sexual conduct.

“It was just one of those situations where it was a huge team effort, I did a small piece of it,” Rouse said. Asked if she was working on another cold case, Rouse said that she couldn’t say. “I feel very strongly that families and victims need closure, so if I can support on the genealogy side to get answers to those families, then I will.”

©2024 The State. Visit at thestate.com. Distributed by Tribune Content Agency, LLC.