Read Caption

Everett Ruess Update: How the DNA Test Went Wrong

Text by David Roberts; Photograph of Utah's Comb Ridge by Dawn Kish

The discovery of explorer Everett Ruess’s skeletal remains, as detailed in the April/May 2009 issue of Adventure ("Finding Everett Ruess"), appeared to be a slam dunk. A team of forensic scientists laid out an impressive case, backed by a DNA test that linked bones found in the Utah desert to the long-lost explorer, an icon of the American Southwest. Well, DNA results are only as good as the process that produces them, and in this case, a peculiar set of blunders managed to complicate, rather than solve, a 75-year-old mystery.

Since he vanished near the Escalante River in 1934, Ruess, a 20-year-old poet and artist, has become a cult hero across the Southwest, famed for his solo journeys across the region which he detailed in rhapsodic prose. His disappearance has turned into one of the most enduring puzzles in the Southwest, one that that countless enthusiasts have tried—and failed—to solve.

An apparent breakthrough came in 2008. Following up on a story originally told by a Navajo man named Aneth Nez, construction worker Denny Bellson came across a skeleton on Comb Ridge in southeastern Utah. The location of the grave seemed to match Nez’s story, which recounted the vicious murder of a young white man in the 1930s. Bellson, along with others, came to believe that the bones must belong to Ruess.

Before the Adventure story was published, Dennis Van Gerven of the University of Colorado performed a thorough forensic analysis, matching a Dorothea Lange photo of Ruess with a facial reconstruction modeled from the skeleton’s skull fragments and teeth. Van Gerven concluded that there was an almost perfect match. To support the findings, the Ruess family requested that a DNA analysis to be performed by Van Gerven’s C.U. colleague, geneticist Kenneth Krauter. Krauter used a technology called the Affymetrix GeneChip, which managed to isolate and compare 600,000 markers obtained from a cross section of the femur with saliva from Everett’s four nieces and nephews. (A typical forensic test compares just 18 markers.) After double-checking the results, the lab sent them to an independent lab for confirmation. In the end, the DNA showed a 25 percent overlap, the exact correlation one would expect between an uncle and his nieces and nephews.

“The combination of the forensic and genetic analyses makes it an open and shut case,” Krauter said last April. “I believe it would hold up in any court in the country.” The bones were shipped to the Ruess family, who planned to scatter the ashes over the Pacific Ocean.

But persistent doubts on the part of bystanders convinced the family to seek a second DNA test to validate Krauter’s result. The lab chosen by the family, Maryland-based Armed Forces DNA Identification Laboratory (AFDIL), has vast experience in identifying fallen soldiers.

In October 2009, AFDIL announced a result that confounded everyone: the skeleton was not only not Everett’s, the lab concluded, but had to be that of a Native American.

The Miscalculation

How could such a drastic mistake be made? The error was not due to sloppy work or contamination, but rather was the result of Krauter’s decision to use relatively new technology in an unprecidented way. The Affymetrix GeneChip technology is widely considered the industry’s gold standard for medical DNA analysis, but remains unproven for forensic work. Krauter, who has worked to characterize DNA samples for more than 20 years, saw the Ruess case as an opportunity to break new ground, and was encouraged by Affymetrix to do so. Unbeknownst to Krauter and his team, however, Affymetrix’s software can produce a false reading—what Krauter calls “noise”—when overly small amounts of DNA are used in tests.

This was vexing for two reasons. In retrospect, Krauter found that even though his team extracted relatively little DNA from the sample, the Affymetrix software produced a result that appeared to be legitimate gene identification data. Even more disquieting, when the computer analyzed the bone samples alongside the data from the Ruess clan, it, upon review, appeared to bias the results in favor of the Ruess family members, yielding a partial similarity between the noise and the family’s DNA at a frequency of 25 percent—the expected value for the test. “We screwed up by relying on the technology too much,” says Krauter. “Fortunately, the error uncovered how the extreme sensitivity can be misleading if a researcher takes its output at face value. We will definitely re-examine how that software can be optimized, and when alternative methods should be used.”

The Fallout

Van Gerven, who performed the facial reconstruction, was dumbfounded by the new results. “I will go to my grave believing that we could not exclude [the match] based on the best anatomical evidence,” he says. “A random skeleton was found that by chance alone matched sex, age, and stature. That in itself is remarkable.”

The upshot of this stunning reversal is that the Ruess family has returned the remains to the Navajo Nation. Ron Maldonado, supervisory archaeologist to the Navajo Nation, is in charge of resecuring the bones in a safe and secret location.

Even as he announced that the family accepted the new conclusion, Everett’s nephew, Brian Ruess, confided that he believed Aneth Nez’s dark tale about witnessing the murder of a young white man in Chinle Wash. He also believes that the story most likely detailed Everett’s demise—it was just that, Brian mused, Bellson had discovered the wrong body. “Everett,” Brian says, “just doesn’t want to be found.”