Twenty-two-year-old Angela Brosso and 17-year-old Melanie Bernas were murdered while riding their bikes in 1992 and 1993, respectively. Photo: Phoenix Police Department

The ancestors of Bryan Patrick Miller may have—accidentally—turned him over to the police for murder.

Last year, Phoenix police arrested Miller for the infamous Phoenix “canal killings” of the early 1990s—with the big break in the cold case finally coming courtesy of forensic genealogist Colleen Fitzpatrick.

In two separate crimes, 22-year-old Angela Brosso and 17-year-old Melanie Bernas were murdered while riding their bikes. Police would find Brosso’s head in the Arizona Canal days after her body was discovered, only to return to the canal a mere 10 months later to find Bernas’ body the day after she went missing. At the time, Phoenix police confirmed that the two murders were connected by a single suspect’s DNA.

However, the trail went cold until detectives with the Phoenix Police Department met Fitzpatrick at the 2014 International Symposium on Human Identification (ISHI). They mentioned the canal killings cold case to Fitzpatrick, who thought she might be able to help.

A few weeks later, the Phoenix Police Department sent Fitzpatrick a Y-STR profile to analyze. A Y-STR is a short tandem repeat (STR) on the Y-chromosome taken specifically from the male Y chromosome.

“My research strategy was to compare the Y-profile provided to me by the Phoenix Police Department with the genetic genealogy databases that are posted online on public websites,” Fitzpatrick explained to Forensic Magazine. “I have developed [proprietary] search strategies and techniques to mine these databases for matches.”

The databases Fitzpatrick is referring to are all public Y-DNA databases posted online by the genetic genealogy community. She estimates there are up to 10,000 such databases currently online, containing somewhere between 300,000 and 400,000 Y-profiles. In Fitzpatrick’s line of work, these databases are essentially a gold mine.

“The 17 Y-Filer loci that law enforcement use for forensic identification are the same that the genetic genealogy community uses for researching family pedigrees,” said Fitzpatrick. “Therefore, Y-DNA results from a forensic lab can be cross-referenced to genetic genealogy databases to determine a possible last name for an unknown assailant.”

In this case, within just a few weeks, Fitzpatrick was able to single out the surname “Miller” as one possibility associated with the DNA profile.

According to KPNX, Phoenix police reviewed records of previous contacts related to the case. They identified Bryan Patrick Miller as someone on a long list of potential leads. The police then independently collected DNA evidence from Miller and matched it against the suspect DNA. The match was positive, and Miller was arrested—25 years after the murders.

Miller has pleaded not guilty. He is scheduled to go on trial in April 2017 and if found guilty, he could face the death penalty.

However, forensic genealogy is an up-and-coming discipline, and the method Fitzpatrick used to initially identify the Miller family name is one with little precedent.

“Whether something is admissible in court is a legal question that is outside of my domain,” Fitzpatrick said. “I will say only that an arrest cannot be made solely on the basis of the data I produce for [law enforcement]. There is much more to the story.”

Still, as forensic technology continues to advance, forensic genealogy may be something both law enforcement and the courts see more and more of.

In terms of Fitzpatrick’s specific research method, it is obviously useful for providing leads on cold cases in the absence of a CODIS hit. But there are other ways to employ the technique as well, according to Fitzpatrick.

“It is also valuable in the case of a mixture of male and female DNA where it is not possible to separate the DNA of the two individuals. Only the male has Y-DNA,” she said. “The technique can also be used by law enforcement to identify John Does.”