Determining the type of biological material (body fluid) left at a crime scene can provide critical context in forensic cases, help establish the relevance of the trace to the alleged crime, and assist in crime case reconstruction. Routine body fluid identification in forensic casework typically relies on presumptive tests based on chemical reactions, immunological assays, enzyme activity, and microscopy. While these methods are well-established and widely used, they can, in some situations, present limitations in terms of sensitivity or specificity.

RNA profiling is an alternative method for body fluid identification. Gene transcription is highly tissue-specific, why RNA expression profiles differ across cell types and may be used to indicate the presence of a particular body fluid. Furthermore, RNA fragments are copies of parts of the genome and therefore retain genetic information, enabling additional analyses simultaneously, including human identification and mixture deconvolution. 

Shotgun RNA sequencing enables comprehensive analysis of all RNA molecules present in a sample (total RNA), including human, microbial, and environmental RNA. In this work, we collected samples (n=298) from a range of forensically relevant body fluids, namely whole blood, dried bloodstains, menstrual blood, vaginal secretions, semen, and saliva. Total RNA was extracted and analysed with shotgun RNA sequencing using the NovaSeq or MiSeq FGx platforms. The human transcriptome and microbiome were characterised for each sample and tissue type. Mixtures were simulated in silico to demonstrate the method's applicability for mixture deconvolution. 

Using selected RNA markers, all tested body fluid types were correctly identified, with the microbiome serving as a secondary identification layer for some tissues. SNP profiles from single-source samples were compiled (minor allele frequency ≥ 0.05; distance between SNPs ≥ 30 cM  ; read depth ≥ 20 reads; genotype quality ≥ 20) using the wgsLR framework (https://mikldk.github.io/wgsLR), and donor identification with high discriminatory power was achieved (LR>1.0E+10) in all tissues except for saliva. Additional information about the donor, including sex, could be inferred from the transcriptome. A workflow for mixture deconvolution using tissue specific mRNAs and results from successful deconvolution of two- and three-person mixtures comprising multiple tissue types and with different contributor ratios will be presented.

Shotgun RNA sequencing provides an alternative method for analysis of the genetic material in forensic samples, allowing for simultaneous body fluid characterisation, donor identification, and contributor assignment.

Authors

• Claus Børsting (Section of Forensic Genetics, Department of Forensic Medicine, University of Copenhagen, Denmark)
• Alberte Honoré Jepsen (Section of Forensic Genetics, Department of Forensic Medicine, University of Copenhagen, Denmark)
• Stine Bøttcher Jacobsen (Section of Forensic Genetics, Department of Forensic Medicine, University of Copenhagen, Denmark)
• Marie-Louise Kampmann (Section of Forensic Genetics, Department of Forensic Medicine, University of Copenhagen, Denmark)
• Jeppe Dyrberg Andersen (Section of Forensic Genetics, Department of Forensic Medicine, University of Copenhagen, Denmark)