It’s been a long journey for Michael Usry Jr. and his family.
Briefly investigated in the 1996 murder of Angie Dodge two decades after the fact and based solely on a partial DNA match to his father, Usry’s life has been haunted by the case for nearly three years.
During that time he’s gone from being considered a suspect — or at least closely related to the suspect — in a murder case, based on one DNA test, to having his entire family cleared of suspicion by another more extensive DNA test.
The story about how he became entwined in a cold case murder investigation 1,900 miles from his home raises alarms about the potential misuse of DNA in criminal investigations and court proceedings, which could lead to false convictions.
A series of unfortunate coincidences
In December 2014, Detective James Hoffman filed an affidavit indicating that the Idaho Falls Police Department had submitted DNA samples from the crime scene to Sorenson Genomics. The company also offered a public DNA database for genealogical purposes in addition to genetic testing. Hoffman ran the DNA submitted as evidence against Sorenson’s genealogical DNA database, which was then public.
Hoffman wrote results showed a close but partial match to a Michael Usry Sr., who had voluntarily submitted DNA to a genealogical database years earlier through a project sponsored by The Church of Jesus Christ of Latter-day Saints.
The affidavit outlined Hoffman’s reasons for wanting to collect DNA from Usry Jr. He produced graphic horror films, and a documentary about people who collect paraphernalia related to serial killers. His name was Michael, matching one of the numerous names Christopher Tapp had given under interrogation for the third man who was supposed to have committed the crime (after police suggested that there had been a third person). Usry also had Facebook friends with family in Rigby.
So on a winter day the New Orleans filmmaker was suddenly pulled into a police station, questioned about a murder committed 20 years ago hundreds of miles away and forced to turn his DNA over to police.
When police tested that DNA, he wasn’t a match. But because of the genealogical testing, Usry’s family came under scrutiny for possible involvement in the crime. Then, earlier this month, the IFPD announced that further testing performed by Parabon NanoLabs indicated it is highly improbable that anyone in Usry’s family is the killer.
It was a jarring contradiction.
Even once Usry himself was cleared, a cloud hung over his family, he said.
If you Google his name, most of the entries are about the Dodge murder. The filmmaker’s IMDb profile isn’t the first entry that pops up. Instead, it’s a news story about him being cleared of the Dodge murder.
Usry had spent almost three years agonizing over his potential relation to the killer, he said in an interview. After he was detained and his DNA was forcibly taken, he decided to make a documentary about the case. He got to know Carol Dodge, Angie’s mother. He promised to do anything he could to help her find the killer.
And since DNA testing seemed to indicate that he might be somehow related to the killer, Usry started having hard conversations with members of his own family. Should they all start submitting DNA in an effort to find out who killed Angie? If they did that, wouldn’t it subject their entire family to greater risk of being pulled into a police station out of the blue?
Usry said before the Parabon results were released, he paid to have his DNA tested further in the hopes that the results could somehow produce a lead on who killed Angie. He did this despite the obvious fears he has about having his DNA in a database, both for himself and for his family.
How is it that one genetic test could say that Usry and the killer are likely to be family, while another says they’re unrelated. The answer has a lot to do with how you define “family.”
What is a family?
The Book of Genesis holds that all humans descend from Adam and Eve. That’s not quite how geneticists see the situation, but they have traced DNA back to demonstrate that there were what they call a “genetic Adam and Eve.”
In short, all humans are part of the same genetic family. The real question isn’t whether any two people are related, but how many generations you have to go back to find a common ancestor.
Police say that in the Sorenson Y-STR test, Usry’s DNA matches 34 out of 35 loci tested (police have refused to release a copy of the report).
The original affidavit used to obtain a search warrant to seize Usry’s DNA indicated that Hoffman had an email exchange with Boise State University geneticist Greg Hampikian, head of the Idaho Innocence Project, who said that the partial match to Usry Sr. meant that he might share a common male ancestor with the killer within three or four generations.
But in a 2015 interview with the Post Register, immediately after news of Usry’s experience had broken, Hampikian said it’s unknown how many generations back you would likely find a common male ancestor between Usry and the killer based on the Y-STR results. It could be five, he said, or it could be 10 generations or more.
The population of men currently living who share a common male ancestor with the killer within 10 generations (the common ancestor might have lived 200 years ago or more) could easily number into the hundreds or thousands, as the Post Register reported at the time. They’re on the same genetic family tree, but they aren’t what we would commonly think of as a family.
What the tests say
The DNA tests performed by Sorenson and those performed by Parabon are different in several ways. Most importantly, the Sorenson test looked only at the Y chromosome, which is found only in men and passed down from father to son. The Parabon test looked at the other 22 human chromosomes, which are generally inherited about 50 percent from the father and 50 percent from the mother.
Parabon’s test, which examines how much overall genetic code Usry and the killer share, found that not only is Usry not closely related to the killer but that his overall genetic code isn’t remotely close to the killer’s.
Parabon’s results indicate there’s at least an 88 percent chance that the killer isn’t related to Usry within six degrees of separation. That means not only that it’s extremely improbable that anyone we would normally consider his family is the killer, but that all of his second cousins once removed are highly unlikely to be the killer as well, as is everyone more closely related to him.
The “unrelated” category is at the very bottom of Parabon’s hierarchy of results. It’s the cutoff where the company won’t even make guesses as to how closely related people are anymore because it’s predictions are no longer reliable. The probability of falsely pegging two people as related is too high.
An explanation of the tests released by Parabon, which said it couldn’t comment on specific cases, indicates that the “unrelated” designation means Usry shares less than 1.6 percent of his relevant DNA with the killer. That’s the upper bound. It could be far less.
No two humans, apart from identical twins, are ever thought to have identical DNA. And that means each human can be uniquely identified by their DNA.
But, Hampikian pointed out that almost no DNA test measures each and every part of the DNA. Rather, small pieces of DNA are measured, and geneticists use statistical laws to show it’s highly unlikely that a match could be false.
Hampikian said Usry’s experience raises several questions worth asking, both about the use of genealogical databases by police and about the use of genetic matches in court. Though Usry shares almost no DNA with the killer, the kind of test performed by Sorenson showed a very high level of match.
Hampikian said Usry’s DNA would have entirely matched the killer’s if a Y-STR test like the one Sorenson completed were performed at 13 loci or 17 loci. Loci are the positions on a chromosome. It isn’t until a 35 loci test that it becomes clear that Usry’s DNA doesn’t match the killer’s, at least with the Y-STR test.
Hampikian said false matches such as this, which show up because tests aren’t sufficiently powerful rather than because two DNA samples actually have the same genetic code, are referred to as “adventitious matches.” And they present real problems.
Imagine if instead of being brought in based on a DNA match, Usry was somehow a suspect in the case, Hampikian said. Then the DNA could be seen as confirmation of guilt, he said, despite the fact that there’s no real match.
“If I had an Innocence Project case like that, I would probably drop it,” he said. “We need to take these coincidental match statistics more seriously. These things are going to happen. These numbers can be misleading.”
Hampikian said he has a forthcoming academic paper exploring the potential risk of adventitious matches in producing wrongful convictions.
And Usry continues to work on his documentary about Angie Dodge’s murder, most recently traveling to Idaho Falls in March when Christopher Tapp, the man convicted in Dodge’s murder, was released from prison after 20 years. Just as was the case with Usry, none of Tapp’s DNA was found at the crime scene.
Today both men are committed to helping Carol Dodge find her daughter’s killer.
Reporter Bryan Clark can be reached at 542-6751.