Groundbreaking DNA technology helped find Martinko murder suspect

Virginia company provides police a 'very scientific list of leads'

Images of a man suspected of murdering Michelle Martinko in 1979 are displayed at a May 16, 2017, news conference at the
Images of a man suspected of murdering Michelle Martinko in 1979 are displayed at a May 16, 2017, news conference at the Cedar Rapids Police Department. Officials presented computer-generated renderings that were created using a database of DNA phenotypes, or the physical expressions of genetic code. Police later turned to the same Virginia DNA company that created the images to help identify a suspect through genetic genealogy. (Rebecca F. Miller/The Gazette)

A burgeoning DNA technique used to arrest the alleged Golden State Killer in April is the same method that helped Cedar Rapids police zero in on identifying a suspect in a 39-year-old cold murder case.

Jerry Lynn Burns, 64, of Manchester,is charged with first-degree murder in the 1979 fatal stabbing of Michelle Martinko, 18.

Martinko, a Kennedy High School senior, was found dead in her family’s 1972 Buick Electra in a parking lot of Westdale Mall. Authorities said there were no signs of robbery or sexual assault, but she had wounds that indicated she fought her attacker and the suspect’s blood was found in the car.

Burns, who was 25 at the time of the crime, apparently had no obvious connection to Martinko and wasn’t on investigators’ radar until earlier this year when they asked a Virginia DNA company for help with the cold case.

“It’s really interesting that Jerry Burns’s name never came up in the investigation,” said Doug Larison, a retired Cedar Rapids police detective who investigated the Martinko case for 10 years ending in 2015.

Parabon NanoLabs, of Reston, Va., is one of the big players in genetic genealogy, helping identify a man accused of sexually assaulting and killing an 8-year-old Indiana girl in 1988 and helping track down a Pennsylvania disc jockey charged with the 1992 murder of a teacher, the Washington Post reported in July.

Several years ago, Parabon used genetic markers from the unknown Martinko murder suspect’s DNA to develop a computer image of the man’s possible look and characteristics — but the image didn’t help identify a suspect, Larison said. But a new online DNA database allowed Parabon earlier this year to home in on suspects, sometimes by name.


Parabon Chief Executive Officer Steven Armentrout wouldn’t talk Thursday specifically about the Martinko case, but told The Gazette how his company uses genetic genealogy to help police agencies solve crimes.

The company extracts genetic markers from DNA supplied by police and uploads the data into GEDMatch, a public genealogy database, Armentrout said. GEDMatch allows people to upload their own genetic data from consumer DNA sites like Ancestry.com and 23andMe and search for relatives.

Because it’s open to the public, law enforcement agencies also may use the site.

“You can upload your file and do a query, asking, ‘Who are my genetic matches?’,” Armentrout said. “If those people have allowed their data to be publicly compared, you’ll get a listing.”

In about 20 percent of Parabon’s 200-or-so cases since May, the company can narrow the list down to one person or siblings, he said. More often, it finds a family tree.

Forensic genealogists use historical records, like obituaries and court documents, to further whittle the list.

“You know your suspect, your unknown, is someone on that list,” Armentrout said. The company provides police “a very scientific list of leads.”

Police then do the on-the-ground work, which might include talking with family members and ruling out suspects, Larison said. But investigators have to be careful to not scare away the person who turns out to be the primary suspect, he said.

Court records released Thursday said police covertly collected DNA from Burns and sent the specimen to the Iowa Division of Criminal Investigation crime lab for analysis. The DCI said Burns was a match to the DNA found in the blood on Martinko’s clothing and consistent with the DNA found on the car’s gear shift.


Larison, who graduated a year ahead of Martinko from Kennedy, said her murder shook him up as a teen and bothered him as it went unsolved year after year. After 10 years in charge of the investigation, he asked to be taken off the case in about 2015, he said.

But Larison often talked with Matt Denlinger, the new officer assigned to the case, before Larison retired in 2017.

“After I retired, I saw the news about the Golden State Killer,” Larison said, referring to the arrest of 72-year-old Joseph James DeAngelo in April in connection with eight California murders. Police believe DeAngelo may have committed at least four other murders and about 50 rapes, according to National Public Radio.

DeAngelo was linked to the crimes through a DNA connection from GEDMatch.

“I immediately called Matt and found out he was already all over it,” Larison said. “He was pushing it (the case) through with the latest technology available.”

Because genetic genealogy is so new, questions remain about how well arrests based on the technology will stand up in court. Earlier this month, John D. Miller pleaded guilty to the 1988 murder of April Tinsley in Fort Wayne, Ind. — a cold case solved with help from Parabon, the Indianapolis Star reported.

Parabon’s Armentrout said developing a family tree of potential suspects through DNA just opens the door on cold cases.

“The detectives will go build a case like they always build a case,” he said. “They will use traditional DNA matching to prove there is a 100 percent match.”

He’s been surprised how many unsolved murder cases don’t involve serial killers. “A surprising number are not. These are people who committed one crime and then just slipped back into society.”

l Comments: (319) 339-3157; erin.jordan@thegazette.com

Give us feedback

We value your trust and work hard to provide fair, accurate coverage. If you have found an error or omission in our reporting, tell us here.

Or if you have a story idea we should look into? Tell us here.