“Next gen sequencing might generate lots of data from small amounts of DNA, but you’re still faced with the same fundamental question — what does it mean?” asks John Butler, special assistant to the director for forensic science at the National Institute of Standards and Technology. “You could detect a single cell on a knife blade, but that doesn’t mean anything — it might have arrived there long before the crime or been transferred there by chance.”
Difficulty with Identifying DNA from Multiple Samples
Another complexity in today’s forensic lab practices occurs when a sample contains more than one person’s DNA. In this case, a scientist must examine the markers, or alleles, at each of the loci.
Imagine three jigsaw puzzles all mixed together with relatively the same sized pieces, each puzzle piece representing an allele. The forensic analyst must not only identify to whom the alleles belong, but also how many people contributed to the sample. This process can be further hindered by a small or degraded sample, which often occurs in tandem, because alleles may be lost in the sorting process or mistakenly assigned. That’s why dealing with mixed samples leads to unavoidable creative interpretation of the data, rather than more traditional scientific methods.
“Ironically, you have a technology that was meant to help eliminate subjectivity in forensics,” Erin Murphy, a law professor at NYU, told The Atlantic. “But when you start to drill down deeper into the way crime laboratories operate today, you see that the subjectivity is still there: Standards vary, training levels vary, quality varies.”
In 2013, the National Institute of Standards and Technology conducted a survey, requesting forensic analysts from over 100 labs to examine a DNA mixture from three people to determine if the suspect’s was included. 24 percent found the data inconclusive, and 70 percent said it might be present. Only 6 percent correctly concluded that the suspect’s DNA was not in the sample.
“It’s the Wild West out there,” Michael Coble, the study’s author and a geneticist explained. “Too much is left to the analysts’ discretion.”
Analysts also differ on how conclusive their results are. Another NIST survey indicates labs interpreting a two-person mixture may report statistical match probabilities that fluctuate by up to 10 orders of magnitude.
“Imagine if you take a pregnancy test and you send it to two different labs,” Hampikian said, “and one said the odds are a billion to one that you’re pregnant, and the other said it’s 50-50.”
Risks of DNA Contamination
These small DNA samples increase the risk of contamination, which can occur in a multitude of ways. The process for replicating small samples into larger ones for testing purposes creates a large risk for contamination itself. With the additional risk of allele dropout during the process, the federal CODIS database—which includes more than 15 million profiles—could potentially implicate someone who was never on the scene.
Other times, a mistake in the collection or storage of the sample by the investigator may cause the contamination. For instance, Adam Scott was wrongfully convicted of rape after his saliva on a plate from an unrelated incident was reused and then found on a genital swab.
In an attempt to remove some of the subjectivity from DNA identification, labs have started to rely on commercial software. For instance, TrueAllele, created by the company Cybergenetics, is an algorithm-based technology intended to detangle mixed DNA profiles and “provide accurate results on previously unsolvable DNA evidence.”
While TrueAllele has already been used in hundreds of cases in the United States, according to experts, Cybergenetics refuses to disclose its source code. Without knowing the assumptions in the technology, it’s more difficult to rely on the software.
To read more, please continue to page 3.