DNA intelligence using sex-chromosome, phenotype-informative and ancestry-informative markers in an Australian population

Single nucleotide polymorphism (SNP) data can be used to infer the biological sex, externally visible characteristics (EVCs) and biogeographical ancestry (BGA) of an unknown individual. There are multiple pipelines available that can be used to generate these inferences and provide investigative leads for law enforcement to pursue. It is important for inference pipelines to be evaluated within a population representative of the intended jurisdiction prior to casework implementation. This study presents the performance of several pipelines using an Australian study population with self-declared biological sex, eye colour, hair colour and recent ancestry. The proportion of consistent results for EVC inference was higher for the HIrisPlex online tool using published interpretation guidelines for eye colour (97%) and hair colour (80%) when compared to the MiSeq FGx® Universal Analysis Software (UAS) for eye colour (74%) and hair colour (69%). For inferring BGA, a principle coordinate analysis pipeline produced the most consistent results when compared to self-declared data (86%). This was improved to 90% when inconclusive results obtained from admixed individuals were analysed with Structure. This study highlights the strengths and limitations of multiple inference pipelines to assist in the development of interpretation and reporting guidelines for Australian applications.