GENES & CELLS
Genetic genealogy could solve crimes
A new way to ID suspects with DNA raises privacy concerns
BY TINA HESMAN SAEY
Within the span of a month, DNA probes
of family trees in a public database helped
police catch two murder suspects.
On May 17, detectives in Washington
arrested William Earl Talbott II, 55, of
Seatac for the 1987 double murder of Jay
Cook and Tanya Van Cuylenborg. A new
DNA sleuthing technique called genetic
genealogy led to Talbott’s capture. His
arrest came just weeks after police in
California used the new approach to
identify a suspect in the Golden State
Killer case (SN Online: 4/29/18).
Arrests in both cold cases are probably
just the beginning of the technique’s use.
Parabon NanoLabs, a DNA-forensics
company in Reston, Va., announced on
May 8 that it has already used 100 genetic
profiles generated from crime-scene
DNA to search the public genealogy
database GEDmatch. One of those profiles led investigators to Talbott. So far,
the company says in about 20 percent of
cases genetic genealogy alone could pick
out a likely suspect.
Another 30 percent of cases may be
solvable with a combination of genetic
genealogy and additional police work,
says genetic genealogist CeCe Moore,
founder of The DNA Detectives, a
genetic genealogy group that helps
adoptees and other people find their bio-
logical families. Moore is working with
Parabon to identify possible perpetra-
tors in murder and rape cases.
Erin Murphy, a professor at New York
University School of Law, acknowledges
the satisfaction of “bringing to justice
people who have evaded responsibility for
pretty heinous crimes.” But she and some
other privacy experts are concerned that
such cases are making innocent people
the subjects of investigations just because
those people happen to share DNA with
a relative whom they may not even know.
Digging through family trees and DNA of
people without their knowledge and consent is akin to searching without a warrant or probable cause, Murphy argues.
What’s more, police might extend
such investigations to less-serious
crimes. “There’s no rule saying police
can only do this sort of genetic sleuthing
if it’s a homicide or rape,” she says. Or,
while looking through financial or other
records of possible suspects from family
trees, detectives might start investigating people for another thing that
they had never been under suspicion
for. “That’s something we haven’t done
in our country. We haven’t said police
should just investigate people randomly
and see what they turn up,” Murphy says.
Betrayed by DNA To help identify a suspect in a murder case, genealogists used crime-scene
DNA to probe a genealogy database called GEDmatch. Two people in the database shared DNA
with the supposed killer (cousins). Reconstructing family trees led police to further investigate and
arrest the suspect (relationships in the family trees are shown relative to the suspect).
In both the Washington and California
cases, investigators created a genetic profile of a suspect from crime-scene DNA
and uploaded the data to GEDmatch.
The database allows people who do DNA
testing through one consumer genetic
testing company like 23andMe to find
DNA matches with relatives who did
testing through a different company (see
Page 14 for more on these companies).
Parabon already had a DNA profile
of the Cook–Van Cuylenborg suspect
handy. It had been used to generate a
sketch of what the suspected killer might
look like at three different ages.
That profile also contained information
that could be used to create the suspect’s
family tree. To build it, Parabon tested
about 800,000 variable spots in the
individual’s DNA, says geneticist Ellen
Greytak, the company’s bioinformatics
director. These spots — single nucleotide
polymorphisms, or SNPs (pronounced
“snips”) — are places in the genome where
people differ in one DNA letter. Scientists
use SNP patterns to detect pieces of DNA
shared by relatives.
Parabon, which was already using
genealogy databases to identify Jane
and John Does, had previously consid-
ered mining the data for clues about
criminals. But, Greytak says, “we weren’t
sure what the public reception was going
to be.” Because the public’s reaction to
the Golden State Killer case was largely
positive, Parabon moved ahead with its
plans, Greytak says. “We now feel that
people are aware, and if they continue
to participate [in GEDmatch], they’re
OK with that usage.”
DNA from Van Cuylenborg’s murder
scene matched two people in GEDmatch,
who both shared amounts of DNA with
the suspect typical of second cousins.
That finding suggested that Moore would
need to go back to at least each person’s
great-grandparents to build the suspect’s
family tree. Moore discovered one branch
of the tree led to Talbott’s mother and
another to his father. Talbott, the cou-
ple’s only son, became the prime suspect.
Police tailed him and got DNA from a cup
he threw away, which matched the DNA
from the crime scene. s