GOLDEN, Colo. – For decades, the evidence in Jeannie Moore's murder sat in a storage vault at the Jefferson County Sheriff's Office.
Leads fizzled out, one after the next. Investigators retired. A DNA sample, taken from the clothes the teenager was wearing when she was killed in 1981, was uploaded to a national criminal database, but no matches were found.
The investigation had gone almost nowhere until May of this year, when Elias Alberti, the lead cold case detective in the sheriff's office, posed an idea to his bosses.
What if the missing link in the investigation – the evidence that would connect the DNA sample on Moore's clothes to the suspect who killed her – was already out there?
What if it was sitting in a public database?
Over the summer, a murder investigation that had gone cold for nearly 38 years was solved in three months: The sheriff's office last week announced that Donald Perea, who died in 2012 at the age of 54, was identified as Moore's killer. She was last seen alive while hitchhiking to work in August 1981. Picnickers found her body five days later in Genesee Park.
"There's 456 months it's been working and literally the case was solved in 90 days, from start to finish," Alberti said. "Finding the suspect and closure in this case happened in weeks. That's how quickly these cases can be solved with this technology. And that's what makes it phenomenal."
Authorities connected Perea to Moore's murder through an emerging type of DNA testing known as genetic genealogy, in which investigators use databases on ancestry websites to identify relatives of a suspect.
Moore's case isn't the only one in Colorado that could benefit from the technology. In June, Colorado Springs police announced an arrest in the 1987 killing of Darlene Krashoc, a Fort Carson soldier. Detectives had investigated Krashoc's killing for years, conducting hundreds of interviews and even creating a composite image of a suspect through a DNA phenotype.
When the detectives submitted the DNA for genetic genealogy testing this year, police said the results pointed to Michael Whyte, a 58-year-old man living in Thornton. Whyte was arrested on a murder charge and his case is pending.
In Jefferson County, investigators are reviewing several other cases to learn if genetic genealogy testing could help, Alberti said. United Data Connect, the forensic DNA company that helped with the Moore case, is working with Metro Denver Crime Stoppers on six additional cases that could benefit from genetic genealogy, said Mitch Morrissey, the former Denver District Attorney who founded the company.
Morrissey said two of those cases "look really promising."
The technology of genetic genealogy is nothing new; genealogy databases and ancestry websites have been around for years, a wealth of knowledge for curious researchers trying to piece together their family tree. But it wasn't until about 18 months ago that investigators began using the databases to solve rapes and murders, in particular cases that had long gone cold.
The first heavily publicized case involving genetic genealogy testing was the arrest of Joseph James DeAngelo, the alleged Golden State Killer, who authorities say is believed to be linked to more than a dozen killings and at least 50 rapes in California dating back to the 1970s.
DeAngelo, a former police officer, was living in the Sacramento area when he was arrested in April 2018. Investigators had plugged DNA from one of the decades-old crime scenes into a genealogy website and used a pool of DNA matches to build a family tree for the unknown suspect. When they narrowed their search to DeAngelo, they took a piece of his DNA from a tissue in the trash outside his home. The DNA matched the DNA found at the old crime scene.
Other cases of genetic genealogy testing have popped up over the last year, including the first conviction involving the use of the technology this summer. A Washington state jury found William Earl Talbott guilty in the killings of a Canadian couple in 1987.
A case that went nowhere
Alberti, a homicide investigator with the Jefferson County Sheriff's Office since 2014, switched to cold cases about two years ago.
A cold case investigation, Alberti learned, involves making calls to people who often aren't around or don't want to talk and reviewing evidence of which there often isn't enough.
The Moore case, in many ways, was no different than the other open homicides on Alberti's desk.
On the day Moore disappeared, the teenager was hitchhiking to her job at a convenience store in Lakewood. Two witnesses saw a vehicle stop alongside the road. Moore pulled on the passenger door, the witnesses said, but it wouldn't open. So the driver reached over and let her inside. The car, described as either a red Ford LTD or Galaxy, pulled away. It was the last time Moore was seen alive.
After her body was found, an autopsy showed that she was killed by several blows to the head.
Other than the witness accounts of her disappearance, few solid leads surfaced.
A semen sample was recovered from a piece of Moore's clothing, but DNA analysis wasn't advanced and the evidence sat in the case file. The murder case went cold.
In 2008, a Jefferson County detective began looking through the case again. DNA technology had advanced, and the semen sample from Moore's murder had been preserved, kept safely in the evidence vault at the sheriff's office.
Over the next several years, investigators re-examined the case and re-processed the evidence, developing a DNA profile from the semen sample. In 2011, the sheriff's office submitted the DNA profile to the FBI's national criminal database, the Combined DNA Index System, known as CODIS.
CODIS, formed in 1990, contains DNA samples of convicted criminals. If the suspect in Moore's murder had a DNA sample in CODIS, the thinking went, it would match with the DNA from the crime scene.
No match was found.
The investigators could only wait for Moore's killer to commit a new crime and find his way into the CODIS database. The case stayed cold.
Another look, and a breakthrough
Little had changed until this year, when Alberti kept seeing the news about genetic genealogy testing. He wondered if it could help crack any of the 37 cold cases in Jefferson County.
Moore's murder "kind of rose to the top," Alberti said.
The case had something that can be difficult to find in decades-old crimes – a "clean" DNA sample in the evidence, Alberti said.
Not only was there a good amount of the DNA, it wasn't mixed with other DNA or substances.
"It was a no-brainer," Alberti said. "We had no choice but to try to get in on it and see if we can use this technology to our benefit."
In June, Alberti's team sent the DNA sample to United Data Connect.
United Data Connect had agreed to a contract with Denver Metro Crime Stoppers involving cases that might qualify for genetic genealogy testing. Crime Stoppers would help fund the testing – each case costs about $5,000 – and United Data Connect would handle the testing process.
In the Moore case, United Data Connect sent the DNA to a lab in Oklahoma, which developed a single nucleotide polymorphisms, or SNPs, profile. A SNP, referred to as a "snip," is the type of DNA profile used in genetic testing and on programs such as Ancestry.com and 23 And Me.
Those two sites don't allow law enforcement to access their databases. Ancestry services Family Tree DNA and GED Match do.
When the investigators received the DNA "snip" profile in late July, United Data Connect uploaded it to Family Tree DNA.
The database returned a group of possible relatives, meaning someone related to Perea – somewhere down the line – had uploaded their DNA to the database. Joan Busse, United Data Connect's genealogist, began piecing together a family tree connected to the DNA.
"What the database is telling you is these are the people that share the most in common with the person who left the crime scene sample," Morrissey said. "You're comparing it to other people in the database, for common segments of the sequence."
Alberti would not say, specifically, which of Perea's relatives the investigators found in the databases. But he gave an example of how the process worked.
Family Tree DNA returned a list of the top 25 matches to the DNA from Moore's crime sample. If one of the matches was, for instance, a third cousin, the investigators would then comb through family history, birth and death certificates and other records to build a family tree back to the suspect.
"It's your job as the user to go back and figure out how these people are related to you," Alberti said.
Where an amateur genealogist might build a family tree forward, starting with their own DNA, Alberti said, the investigators traced a family tree backward. The results led to Perea's two daughters.
When investigators learned that Perea had died in 2012 and that he was cremated, they needed his daughter to provide a DNA sample. She agreed to help.
"We wouldn't be sitting here having this conversation if it wasn't for her," Alberti said.
The investigators took the daughter's DNA sample and "basically ran a reverse paternity test," Alberti said. Early this month, they received the results: Perea was a match. The DNA showed that he was 3.3 trillion times more likely than any other person to have killed Moore.
"I understood how it was supposed to work," Alberti said. "Still, I was blown away at how fast it went. How quick from start to finish. And that's mind-boggling when you think about a case that's been sitting here for 38 years, and literally in weeks we were able to put closure to it."
Can this work with other cold cases?
More than 1,700 cold case are open in Colorado, according to the Colorado Bureau of Investigation's online records. Likely only a small percentage of those cases would even qualify for genetic genealogy testing, Morrissey said.
"You have to have a lot of DNA, and you have to have, pretty much, a single-source sample," Morrissey said. "You have to have a single source, otherwise you can't sort one out from the other. There are just a lot of cold cases out there where they either don't have enough DNA, or they're a mixture."
When interested investigators ask Morrissey if genetic genealogy can help their case, he sometimes has to turn them away: "We'd love to help, but don't waste your resources."
Still, the former prosecutor called the genealogy testing a significant advancement. The analysis of DNA from crime scenes was the first step. Then came the CODIS database, creating a collection of DNA that those crime-scene samples could be tested against.
"But when the guy isn't in the database, what do you do?" Morrissey said. "Sadly, a lot of people do nothing."
United Data Connect began as a company that specialized in familial DNA testing. It's similar to genetic genealogy but doesn't use as broad of a pool. Only 13 states in the country have familial DNA databases, which collects DNA of convicted felons. If the person who left DNA at a crime scene has an immediate family member – a father, brother, sister or mother – who is in the familial database as a convicted felon, then a match can be found.
But even then, Morrissey said, the pool is limited. Genetic genealogy testing broadened that pool.
Both Morrissey and Alberti cautioned that genealogy results, no matter how big of a breakthrough, are just investigative leads. Even in the Moore case, investigators still needed the DNA sample from Perea's daughter to confirm a match.
The genealogy testing, though, took a case from nowhere to somewhere. There likely would have been no other way to solve the case – Moore and Perea had no prior relationship, and no one in their families knew one or the other.
"The only connection they had was that she was hitchhiking and he stopped to pick her up," Alberti said. "And their last connection is when he killed her."
