The computer model connects family members with the help of genealogical databases and law enforcement forces

The notion of using genetic ancestry databases to solve crimes recently passed from hypothetical to credible was used when police used an online genealogical database to find the so-called Golden State Killer, a serial criminal who had terrorized much of California in the 1970s and 1980s. Now, in a study published Oct. 11 in Cellresearchers report ways to expand this type of survey.

Specifically, they published a calculation method for linking individuals from ancestry databases to those of law enforcement databases. These two databases use completely different genetic marker systems. Investigators report in a proof of principle with 872 individuals that more than 30% of close relatives (siblings or siblings) can be accurately compared to the parent using non-overlapping genetic markers from two different databases.

"There is a problem inherited from the fact that so many DNA profiles have been collected with this ancient genetic marker system used by law enforcement since the 1990s. The system is not designed for queries. more complex that presently have an interest, such as the identification of people, represented in a DNA mixture or identifying the parents of the contributor of a DNA sample, "says senior author Noah Rosenberg, professor of biology at Stanford University. "In this study, we were trying to ask whether a newer, more modern genetic marker system could be tested against the old system and continue to obtain matches and find relatives. "

The database used by the FBI and other law enforcement agencies is known as the Combined DNA Index System (CODIS). It relies on tandem repeat markers (STRs), a type of copy number variation, in non-coding regions of DNA. (The system initially used 13 markers, but has recently been updated and currently has 20.) On the other hand, pedigree databases look for differences in single nucleotide polymorphisms (SNPs) on hundreds of thousands of genome sites.

In a study published last year, the Rosenberg team said that the software could associate individuals appearing in both databases, even with genotypic datasets without shared markers. They corresponded to more than 90% of people using version 13 of CODIS and up to 99% with 20 markers. The key idea is that each STR marker is surrounded by SNPs that are usually inherited with the STR. As a result, the genotypes of a person for these SNPs can partially predict the genotype of the neighboring STR and vice versa. When these subtle correlations are accumulated on many STRs, it becomes possible to match an SNP profile with a STR profile.

The new document builds on this research to determine if the same approach would connect close family members. They found that, when one individual had been analyzed for STR markers and the other for SNP markers, about 30% to 32% of parent-child pairs and 35% to 36% of sibling couples sisters could be linked.

In the Golden State Killer case, law enforcement forces submitted DNA collected on one of the crime scenes for genotyping the SNP and then used a database of Open source pedigree to link this profile to other people present in the database. However, the technique described in the new article suggests that family searches could link CODIS members to family members belonging to an ancestry database or vice versa.

The study was intended to provide data to discuss many issues related to forensic genetics and the confidentiality of genomic information, says Rosenberg. "We wanted to explore how these different types of databases could communicate with each other," he said. "It is important for the public to know that information between these two types of genetic data can be connected, often unexpectedly."

When the current policies regarding DNA evidence were established, it was impossible to establish that link. "We showed that it was possible to extend the investigative scope of STR forensic profiles beyond what we previously believed to be the limit," he adds. -he.

In the paper, researchers point to other policy issues regarding this expanded capacity. For example, some populations are overrepresented in the STR databases of law enforcement. The increased use of database searches could alter the calculation of those accessible to investigators from the profiles of these databases. "There has already been a lot of legal analysis on how STR databases are used," said Rosenberg. "With this study, we suggest that SNP databases and their links to STR databases are also taken into account in this analysis."

The new findings have applications in other areas of study than law enforcement. For example, ecologists studying field organisms could use this approach to determine whether animals living in a given geographic area were descended from animals whose DNA had been collected during a previous campaign. even if only STR data were available for older samples. Binding tools could also potentially be used to bind old human DNA fragments together, for example, when multiple samples are being tested on an old burial site.

Explore further:
Should the police be allowed to use genetic information in public databases to track down criminals?

More information:
CellKim et al. "Statistical detection of relatives typed with disjoint medico-legal and biomedical loci." http://www.cell.com/cell/fulltext/S0092-8674(18)31180-2, DOI: 10.1016 / j.cell.2018.09.008