Democratizing investigative genetic genealogy

The genetic component of genetic genealogy (GG) is an estimate of kinship, often conducted at genome scales between a great number of individuals. The promise of GG is substantial: in concert with genealogical records, it can identify a person of interest without a direct reference sample to compare against. However, there are many caveats and pitfalls to genome-scale assessments of kinship, the most pressing of which is imputation. While genotype imputation is currently used in forensic GG cases, the effect of imputation on basic population processes and forensic samples is poorly understood. The medical genomic literature suggests that kinship is well-estimated from imputed data and that inferring kin is robust to many population genetic processes. For example, properly imputed low-pass shotgun sequencing (a 1× genome) is more accurate than competing microarray technologies while producing genotyping error rates that are less biased in populations such as African Americans. Imputation is also used on archaeological samples (recently, on locks of hair from Ludwig Von Beethoven), suggesting a resilience of the approach to certain kinds of sample insults.

A second concern with GG is accessibility. Solutions like QIAGEN’s Kintelligence Kit are available (though expensive) to some forensic laboratories, but they are far less powerful at identifying distant relatives than whole genome techniques. Forensic laboratories may also generate shotgun sequencing data from inexpensive and accessible third party facilities, however doing so risks breaking the chain of custody. Further, such facilities are unlikely to provide raw data, which may create issues at trial (e.g., discovery). Imputed low-pass sequencing provides an alternative solution that enables accurate genome-scale kinship estimation using inexpensive desktop sequencers.

We propose to examine whether genotype imputation, when conducted on the appropriate chemistry coupled with the appropriate bioinformatic preprocessing, provides a way to assess kinship in-house at affordable prices. Pilot data demonstrate that imputed shotgun sequencing from a desktop sequencer (the MiSeq;Qiagen) generates genome-scale data sufficient to identify distant relatives (up to 3rd cousins). Further, large-throughput instruments (e.g., the NovaSeq; Illumina) may be able to provide the equivalent output at prices comparable to microarrays (~$100/sample) with improved results (both more accurate and with less population-specific bias). We will develop a reproducible end-to-end scientific workflow for conducting imputation on forensic samples. Further, an analysis of the resiliency of this system to forensic and population genetic scenarios will be shared with the scientific community. CA/NCF