Det. Steve Smith, lead cold case detective in the Christine Jessop case. It was early this summer when detectives to realized that an experimental investigative technique might be leading to a break in one of Ontarios highest profile cold cases. From a len
Toronto Police Service detective Steve Smith is working on about 70 cold cases—using DNA databases photographs courtesy of toronto police service

This Cop is Cracking Cold Cases With DNA

How genetic genealogy is solving murders from decades past
Rosemary Counter

June 20, 2024

The rise of consumer genealogy websites—think Ancestry.ca or 23andMe.com—has created a surprising new industry: investigative genetic genealogy, or IGG. That’s when databases of DNA samples from regular citizens are used alongside genetic evidence from cold-case files to solve seemingly unsolvable crimes from decades past. Leading the exciting new field is Toronto Police Service Detective Sergeant Steve Smith, who’s spent the last five years becoming a genetic genealogy expert, thanks to a $1.5-million grant from the provincial government in 2022. 

Recently, Smith’s team used IGG to identify and arrest the suspect of two 1983 murders in Toronto. The Niagara Regional Police Service also used the tech to locate a suspect in the 1999 murder of a 26-year-old Torontonian. Here, Smith talks about how cold cases became his passion, new developments in the rapidly evolving field, how much it costs to solve a cold case and, of the 70 or so files he’s currently working on, the one he’d most like to solve. 

You have the absolute coolest job. How did you land there?

I was working in bank robberies when I was brought to homicide on a special project. After, I started working on some cold cases, and the unit commander asked if I’d be interested in taking on cold cases full-time. 

Around 2019, I went to the Canadian Police College for the first ever Unsolved Historical Homicide course. They brought in the investigators who found the Golden State Killer from Southern California, using DNA evidence they uploaded to GEDmatch. Even though the course was really technical and I didn’t know what they were talking about, I was hooked. 

Can you explain the technology to me in layman’s terms? 

When we recover DNA from a crime scene, we send it to the Centre of Forensic Science to create an STR, or short tandem repeat, a DNA profile. This is basically the profile you’d take to court, with 21 DNA markers. STRs give you unknown male or female profiles. We upload them to our national database but often have no hits. 

This is where investigative genetic genealogy comes in. From there, we take a sample of the STR extraction and send it to a private lab in the U.S. They put it into a fancy DNA-sequencing machine which generates what’s called an SNP, or a single nucleotide polymorphism. If done well, this profile has hundreds of thousands or even millions of DNA markers—the genes between your genes that show a person’s skin colour, eye colour, hair colour, where their family is originally from. 

Where do the genealogy sites come in? 

We can upload an SNP profile to sites like GEDmatch or FamilyTreeDNA, which both allow police involvement if users agree. Then we’re up and running. We get a list of people on file who match with the offender’s DNA. This could be anywhere between zero and a couple hundred. If we match with a close relative, you could finish a case in 24 hours. If it’s fourth or fifth cousins, you’re looking at six or eight months to solve the case. Our genealogists use open-source data to build the family tree back in time to find the most recent common ancestors and then work down toward the offender. When we’ve narrowed our search down to one family of interest, then we’re looking at more traditional police work.

Which cases did you choose to get started with IGG and why did you choose them? 

We started with two cases where we thought IGG might be useful: the case of nine-year-old Christine Jessop, and the murders of Susan Tice and Erin Gilmour. Both were high-profile murders of vulnerable people—a small child, and two women who were alone in their own homes—with good DNA samples on file from the offenders. 

1983-30

Spoiler alert: You solved both of these. Tell me about one of them?

For the Christine Jessop case, we sent the DNA sample for an SNP and uploaded it to GEDmatch and Family Tree DNA. We were lucky to get one match on the paternal side and another on the maternal side; both of the families had immigrated from the U.K. to the Belleville area. From there, we narrowed it down to two families in Ontario. We identified the offender, but he’d since died by suicide. It’s great to have finally identified him and give the family some closure, but it’s hard to know that he never had to stand trial for the crimes he committed. 

What does all this cost?

The cost-benefit analysis is unbelievable. Think about what it would cost for one or two investigators to work on a case for 40 years. That’s well into the millions of dollars. Each one of these IGG cases costs us about US$8,000 and a few months of investigative time. We can usually solve a case for under $50,000, which includes all the officers, genealogists and testing. 

You got a $1.5-million grant two years ago. Is that enough?

The grant was amazing, but we really need consistent funding. Without it, we won’t be able to continue to employ genealogists. This is the biggest hurdle for us right now. And the fact that we can only use two of the genealogy sites and neither of the big ones, Ancestry.ca and 23andMe.com, which in total have over 40 million DNA profiles. We can’t use anyone’s DNA without their authorization. 

Unless someone’s a serial killer, why wouldn’t they agree? Should Ancestry and 23andMe change their terms of service? 

I think so, but lots of people disagree with me, citing privacy, which was built into these sites’ terms of service before we knew anything about IGG. We recently went in front of the Senate asking to expand our DNA database. We were asking for much less than many countries: just DNA upon conviction of an offence and limited familial testing. Both were voted down—unfortunately, if you ask me. Ninety-nine per cent of the population would call the police if they knew someone in their family had committed a violent crime. To me, this is just like using your genetics to call 911. 

Is there any case in particular that you want to see solved?

All of them, really, but one that sticks out right now is the Margaret McDonald case. It happened in the Leaside area of Toronto in 1994. She was a lady in her 80s, living alone, who was taking a nap when somebody broke into her house. They found her upstairs, savagely assaulted her and then beat her to death in her own bed. It was a brutal attack, and the suspect left DNA all over the place. We figure the offender was pretty young at the time, late teens or early 20s, so he could still be walking around. I really hope we get him. 

How can average people help out? 

You can assist law enforcement by taking your Ancestry profile and upload it to either GEDmatch or FamilyTreeDNA. This is the biggest thing we’re looking for from people. The more people we have, the better chance we have at catching offenders. If everybody on Ancestry uploaded their profile, we’d be able to solve every cold case in the world.