NEW YORK (GenomeWeb) – Drawing on techniques it honed working with ancient DNA, researchers from Stanford University have designed a targeted next-generation sequencing panel for forensic applications that they say is amenable to degraded DNA and mixed samples.
The Stanford team, led by Carlos Bustamante, is the latest to enter the still young but rapidly growing field of NGS-based forensics. Bustamante told GenomeWeb that the idea for a targeted panel came from work the group had been doing on whole-genome capture of ancient DNA, and that it would complement other approaches.
Bustamante's lab previously developed a technique called WISC, for whole-genome in-solution capture, which captures ancient human DNA from samples like fossilized bone, which contain many other contaminants from the environment as well as genetic material from microbes. WISC uses modern genomic DNA to produce biotinylated RNA probes that can grab corresponding stretches of ancient human sequence out of solution, leaving behind sequences from microbes or other environmental contaminants.
The approach is effective, yielding up to 150 times the DNA in a sequencing library compared to standard library prep methods, and Bustamante said that some of the same problems that plague ancient DNA samples, such as contamination and degradation, are also a problem in forensic samples. However, the whole-genome approach is costly and may be overkill for forensic applications, he said.
"For a lot of the questions, you'd like to ask, what you care about are the phenotypically important or ancestry informative SNPs," he said.
So far, he said, the team has designed two versions of the panel, dubbed PhenoCap. The first version covers 24 SNPs indicative of hair and eye color as well as the full mitochondrial genome, while the second version includes the mitochondrial genome, over 3,500 SNPs, and the Y chromosome.
The panel uses the same in-solution capture technique as WISC, which Bustamante said was preferable to strategies like amplicon sequencing, because the DNA is so fragmented — often in pieces just 50 bases long.
During a presentation at last month's Advances in Genome Biology and Technology meeting in Marco Island, Fla., Bustamante said that his group had applied PhenoCap to a sample from Peru, generating 1 million reads. The panel predicted that the sample had an 80 percent probability of having black versus brown hair and a 99 percent probability of having brown eyes.
In addition, researchers from Bustamante's lab, including Meredith Carpenter, who did much of the development on WISC and PhenoCap, have formed a spinout company, IdentifyGenomics, commercialize the technique. The company is still in stealth mode.
Bustamante said that his lab is also collaborating with a law enforcement agency on an unsolved sexual assault and homicide case from the 1990s. The team applied the panel to four DNA samples — two from the perpetrator's sperm, one sample from the victim's pubic hair, and one from the victim's hair.
The work is ongoing, but so far, the NGS approach has found that the perpetrator is likely of Northern European ancestry and has no evidence of admixture.
Currently, forensic profiling is based on STR analysis with capillary electrophoresis or analysis of mitochondrial DNA with PCR and Sanger sequencing.
Bustamante said that his lab's capture panel offers an advantage over those techniques in that it may work better with mixed samples and highly degraded DNA. DNA that is highly degraded in small fragments less than 100bp is not readily amenable to PCR, and the older the sample, the more degraded it likely is, making the PhenoCap technique especially useful for decades-old cold cases, he said.
Conventional tools, like the CODIS markers, are "well-established" and there is a legal precedent for using them in the court, so neither prosecutors nor defense attorneys have a major impetus for bringing in new technology that is not as proven, Bustamante said. But even though there "hasn't yet been a huge adoption of NGS in this area, we believe there's a huge opportunity."
A number of other research groups, as well as both Illumina and Thermo Fisher Scientific's Ion Torrent, think so too.
Earlier this year, Illumina launched a system specifically for forensic applications, the MiSeq FGx system, which includes a library prep kit and analysis tools. Ion Torrent has also commercialized a human identity panel, AmpliSeq HID.
In addition, a number of other researchers in both the US and Europe are looking to move existing methods, such as STR profiling and mitochondrial DNA analysis to NGS platforms, and are also looking to develop SNP-based panels.
Bustamante said that he has not tested the MiSeq FGx system, but has used both Illumina's NextSeq 500 and the Ion Proton for forensic sequencing work.
Besides criminal forensics work, which Bustamante thinks is still far from being implemented in court cases, he sees a huge opportunity in using the techniques on museum samples, for instance. A targeted approach would enable hundreds of samples such as skull collections to be screened at a reasonable cost. "Most museum samples are far from being characterized genetically," he said.
Earlier this week, Bustamante and researchers from the University of Copenhagen, published an analysis in the Proceedings of the National Academy of Sciences of ancient DNA taken from three human samples extracted from a 17th century burial site on the Caribbean Island St. Martin. They used WISC to show that the individuals were of African ancestry and likely victims of the transatlantic slave trade.
Bustamante said that while the approach involved doing whole-genome sequencing of just a few individuals, using the PhenoCap panel could enable similar analyses of hundreds of samples in a cost-effective way.