As the Defense POW/MIA Accounting Agency exhumes the commingled remains of 388 sailors and Marines from the USS Oklahoma who are buried in Hawaii as unknowns, some outside experts are skeptical that DPAA scientists will be able to identify them based on outdated DNA testing methods.
The DPAA lab usually uses mitochondrial and Y-STR DNA testing in its work, methods that could put the Hawaii laboratory and the Armed Forces DNA Identification Laboratory in Delaware at a disadvantage.
“The technology they’re using is basically 1990s technology,” said Cecil Lewis Jr., presidential research professor, associate professor of anthropology at the University of Oklahoma, and co-director of its Laboratories of Molecular Anthropology and Microbiome Research. “Genomic science has aggressively moved past it.”
Ideally, nuclear autosomal DNA testing would give the agency the best chance to identify the crewmembers, who died when the Oklahoma was sunk during the attack on Pearl Harbor on Dec. 7, 1941.
Nuclear autosomal DNA, which comes from mother and father, is like a fingerprint, exclusive to a single individual, and is better suited to making speedy and precise identifications, experts said. Because of its level of certainty, it could stand alone as the only step in a successful identification process.
Mitochondrial DNA comes only from the mother, with Y-STR coming from the father. They are not precise methods and have a margin for error. Experts say they are better used in confirming an identification after weighing circumstantial, dental and other types of evidence, and in excluding a potential match.
A search of the DPAA website indicates that the lab relies almost exclusively on mitochondrial, or mtDNA, testing to identify remains.
Since 2012, DPAA — and its predecessor agency, the Joint POW/MIA Accounting Command — have put out more than 420 news announcements about remains being identified. More than 400 mention mtDNA as the means of identification. Only a few described use of nuclear autosomal DNA.
DPAA and AFDIL officials said they do use nuclear autosomal DNA but declined to provide the number of times it has been attempted and was successful in recent years.
“As far as numbers go, we don’t talk numbers,” said Tim McMahon, the Armed Forces DNA lab’s deputy director for forensic services. “We run it on a case-by-case basis, and the results and the availability of family references will dictate when and how we will use testing.”
DPAA’s reluctance to divulge its success rate is unusual in the scientific community, said Christina Warinner, who works with Lewis at the University of Oklahoma lab.
“It is considered field standard for ancient DNA labs to publish their success rates for each study,” she said.
Which test is best?
Last year, AFDIL and the Defense POW/Missing Personnel Office stated that nuclear autosomal DNA could be used only on current war casualties.
“Unfortunately, nucDNA is not a viable tool in older remains due to many environmental factors that cause the nucDNA to degrade,” the DPMO website said before the agency was wrapped into the newly formed DPAA.
Another problem often cited by DPAA and AFDIL officials is the need for immediate family members to make a nuclear match.
“What people don’t understand is that autosomal DNA, because you receive half from your mother and half from your father, in successive generations those autosomal-STR markers are diluted, and we may have very few relevant autosomal markers present in the family reference samples that we have — if we even have a viable autosomal-STR family reference sample available,” said Navy Capt. Edward Reedy, new DPAA medical examiner. Sometimes, there are not enough family members available for comparison testing, or the sample could be too diluted to use. “So to sum up, generally we can’t make an identification based solely on autosomal DNA or Y-DNA because we don’t have enough autosomal markers to compare against and we may not have the proper family reference samples.”
Experts contacted by Stars and Stripes said that obtaining viable nuclear autosomal samples is more difficult than mtDNA, and that inexperienced analysts often have a hard time getting a complete DNA profile from a sample because of environmental factors, which could include the acidity of the soil or burial of remains with metal objects. However, experience and setting the proper extraction parameters have proven effective in the private sector in recent years.
Ed Huffine, former vice president of international development for forensic DNA firm Bode Technology Group Inc., used nuclear autosomal DNA last year to identify the remains of Army Pfc. Lawrence Gordon, who was killed in 1944. Huffine, who worked for AFDIL, said nuclear DNA testing enabled him to reach beyond immediate family to nieces and nephews and even cousins to make the identification.
An internal government email obtained by Stars and Stripes seems to acknowledge the AFDIL lab’s shortcomings.
“Obviously the Bode lab is superior to AFDIL, as we should have been pursuing [nuclear] DNA all this time,” said correspondence between JPAC personnel and officials from the Armed Forces Medical Examiner’s office, which includes AFDIL. The name of the sender and recipient were redacted due to government privacy policies.
Damir Marjanovic, a professor at the University of Sarajevo and former forensic genetics expert for the International Commission on Missing Persons tasked with identifying war victims in Bosnia, said that nuclear DNA is the No. 1 choice when possible.
“We have successfully used autosomal nuclear DNA in identification of the WWII victims from the mass graves in Slovenia and Bosnia-Herzegovina,” he said. “We have obtained useful autosomal nuclear DNA profiles for more than 90 percent of analyzed skeletal remains.”
Marjanovic said that he has obtained useful autosomal profiles from skeletal remains from the 13th and 15th centuries.
He would not comment on the U.S. government’s capabilities.
Joshua Hyman, director of the University of Wisconsin’s Biotechnology Center DNA Sequencing Facility, said that the debate over identifying remains shouldn’t focus just on mitochondrial versus nuclear testing. Instead, the government should work on getting its labs up to speed with emerging technologies and using DNA as part of the initial analysis of the evidence and not as a last step in identification.
He said these dated practices are happening in forensic labs across the country, including those run by the U.S. government.
Hyman and his colleagues at the University of Wisconsin, as well as Lewis and Warinner, said that next-generation sequencing is the way forward. They have been using that technology for years.
Meanwhile, the Defense Department labs have experimented with the technology but have not perfected or adopted it, according to documents.
Huffine said he worries that family members will die while waiting for DOD to identify Oklahoma remains.
“It’s like they’re using a 20-year-old computer,” he said. “Change needs to come if we’re going to ID our missing in a reasonable time frame.”