Jianye Ge estimates that over 15 million criminal cases and missing persons identifications have used familial testing since it skyrocketed in popularity a few years ago, and especially after the Golden State Killer case.
Ge, the associate director of the Center for Human Identification at The University of North Texas Health Science Center at Fort Worth, is thrilled the technique has been so helpful, but is cautious of the potential for errors. In a study titled "How Many Familial Relationship Testing Results Could Be Wrong?", published in PLOS Genetics, Ge reviewed worldwide familial practices, ultimately concluding STR-based technology has limitations and is more prone to errors than SNP-based technology.
"We used a realistic model to estimate that tens of thousands of cases could be wrongly interpreted over the past two decades,” Ge told Forensic. “Many in the relationship testing community understand the limitations of the current STR technology, but no one estimated how significant the problems were. Our study is the first study to quantify the impact of the technology limitations,” Ge told Forensic.
Using a 15 autosomal marker kit and a likelihood ratio of 100 for inclusion, Ge and colleagues used an extrapolation model to calculate a false interpretation rate of 0.058% for trio cases (biological mother, child, and alleged father), 1.14% for parent-child cases, and 20.82% for full-sibling cases. Kits with 21 STRs had significantly reduced false rates. The trio cases, with a threshold of 100, yield false negative and positive rates of 1 in 111,000 and 1 in 10 million, respectively. Overall, the study authors conclude, the current STR marker systems do not address the false positive problem, nor are they more genetically informed.
Thus, Ge encourages the forensic community to move toward the use of high-density SNP panels.
“Traditional STR-based technology only tests STR panels, which usually contain 20 to 25 STRs, while high-density SNP-based technology tests hundreds of thousands to millions of SNPs. Even though the SNPs have lower powers of discrimination per locus basis compared with STRs, the cumulative powers of discrimination with so many SNPs are much higher than the current STR panels,” Ge said.
In his paper, Ge provides actionable recommendations for labs wanting to implement high-density SNP testing. The first step, he told Forensic, is for labs to investigate different high-density SNP technologies and run a cost benefit analysis to decide if they should bring the technology in-house or outsource to service providers.
The development of a standard workflow and software tools is also important, as well as educating users on the strengths and limitations of the technology. Validation of SNP-based methods, however, is the most critical aspect.
“With solid validation, high-density SNP panel-based technologies could meet scientific scrutiny as well as Frye or Daubert admissibility standards, in a similar fashion to that of current forensic DNA testing systems,” Ge writes in his paper.