Genetic genealogy: a family-tree lead, then a direct DNA comparison
Genetic genealogy helped investigators reach Joseph James DeAngelo through people who shared ancestry with him. It did not begin with his name attached to a matching offender profile. NIST forensic scientist John Butler's explanation of the Golden State Killer investigation separates the search for relatives from the later collection of confirmation samples. DNA had already connected crimes, but investigators still needed to identify the person behind the profile. The breakthrough combined a preserved biological sample, a genealogy database, family-tree research and a direct comparison with DNA obtained from DeAngelo himself. Each stage supplied something the preceding stage could not establish alone.
The National Institute of Justice's funding record documents groundwork that preceded genetic genealogy. Its Solving Cold Cases with DNA program supported five Ventura County awards between 2005 and 2014, two awards to Orange County, two to Sacramento's district attorney and one to Placer County. NIJ describes the work as connecting DNA from sexual assaults and homicides to one person operating under several different names assigned by investigators. This was already a major evidentiary connection. It joined crimes across jurisdictions without necessarily identifying their common perpetrator. The grant history places the later breakthrough within years of laboratory work, preserved evidence and cooperation among separate agencies.
The technical difference is set out in the Justice Department's 2019 interim policy on forensic genetic genealogical searching. Conventional forensic comparisons use short tandem repeat markers, or STRs. Genetic genealogy instead examines a dense set of single nucleotide polymorphisms, or SNPs, to identify shared stretches of DNA that can indicate relationships. Larger inherited blocks generally suggest a closer relationship; smaller blocks can connect more distant relatives. The policy describes comparing those profiles through genealogy services whose users have submitted samples or uploaded data. These were different markers, algorithms and databases from the conventional CODIS system used in criminal investigations.
Butler's explanation identifies the practical requirements in the 2018 case. Investigators needed a sample of sufficient quality and quantity for the more extensive profiling; he identifies an old semen stain from a sexual-assault kit. They also needed enough people in a genealogy database to make a useful relative match possible. GEDmatch provided the shared resource. The resulting genetic genealogy leads still demanded extensive detective work to construct family trees. Butler explicitly states that he did not participate in the investigation. His explanation is an official specialist's later description of the method, rather than a claim to have personally made the family-tree connection.
A separate 2018 study by Yaniv Erlich and colleagues tested the power of long-range relative searching using data from 1.28 million consumer-genetics customers. Nearly 60% of searches returned a match sharing enough DNA to suggest a third cousin or closer. About three-quarters of the sample had primarily Northern European ancestry, so the findings did not represent every population equally. The study also modeled how geography, age and sex could narrow a family-based candidate group. These were research results about genetic genealogy's capabilities, rather than measurements of DeAngelo's particular family tree. They explain why a distant relative match can be useful while still leaving many possible individuals to investigate.
Erlich's team emphasized the accessibility and accuracy of genealogical records and the time required to organize them. The reconstruction of DeAngelo's identification likewise describes investigators building trees from distant relatives and eliminating candidates using age, sex and residence. A family connection establishes a set of possibilities, rather than identifying which relative deposited biological material at a crime scene. Genetic genealogy can extend the search beyond people already represented in an offender database, but the paper trail has to connect generations correctly. A mistaken relationship or an unsuitable candidate still has to be detected through further investigation and, ultimately, direct testing.
NIJ records the decisive confirmation separately: police obtained a biological sample from material DeAngelo had discarded and matched it to crime-scene DNA before his arrest on 24 April 2018. The historical chronology also describes samples from a car door handle and a tissue recovered from his rubbish. Those comparisons tested DeAngelo himself rather than a distant relative's genetic resemblance. His guilty pleas followed on 29 June 2020, covering 13 murders and 13 kidnappings; he received life imprisonment without parole that August. Genetic genealogy generated the route toward a suspect, while the direct DNA comparisons and subsequent criminal proceedings supplied different forms of confirmation and legal resolution.
The later federal policy also limited when genetic genealogy could be pursued. Approved in September 2019 and effective that November, it postdated DeAngelo's arrest. It required an unsuccessful CODIS search and pursuit of reasonable investigative leads before the technique was considered. A genealogy association supplied a lead, and the policy expressly barred arrest based solely on it. Direct STR comparison with the suspect was required. These were instructions for investigations covered by the new federal policy, rather than a description of rules governing every step taken in California in 2018. The policy formalized distinct standards for generating a lead and confirming a suspect's DNA.
- John Butler, NIST, National DNA Day and the Birth of Investigative Genetic Genealogy, 25 April 2019 — www.nist.gov
- National Institute of Justice, Serial Killer Connections Through Cold Cases, DeAngelo grant and identification records — nij.ojp.gov
- U.S. Department of Justice, Interim Policy: Forensic Genetic Genealogical DNA Analysis and Searching, 2019, sections III–V — www.justice.gov
- Yaniv Erlich et al., Identity inference of genomic data using long-range familial searches, Science, 2018 — pmc.ncbi.nlm.nih.gov
- Wikipedia, Joseph James DeAngelo, identification and court chronology — en.wikipedia.org