One of the main challenges in forensic genetics is the analysis of DNA from mixed biological samples, where the presence of multiple contributors complicates STR profile interpretation. Despite advances in probabilistic interpretation software, mixture deconvolution remains limited by factors such as overlapping alleles, similar contributor ratios, and PCR artifacts including drop-out and drop-in.

Single-cell DNA typing has emerged as a promising approach, as it enables the physical separation of individual contributors prior to amplification and genotyping, thereby facilitating mixture deconvolution. However, most current methods rely on complex, time-consuming, and costly micromanipulation techniques, which limit their applicability in routine forensic workflows.

Here, we present a simplified single-cell typing workflow based on the use of the DispenCell System, which uses impedance-based detection to enable rapid and automated isolation and dispensing of individual cells into separate wells within seconds. The isolated cells are subsequently processed using direct PCR, consistent with established approaches for low-template DNA analysis.

Proof-of-concept experiments yielded positive results using the HEK293 human cell line, as well as single-source blood and saliva samples. Other body fluids of forensic interest, such as vaginal secretions, menstrual blood, and sperm, have not yet been evaluated. The workflow will be assessed using mock mixtures of biological fluids with increasing complexity in terms of contributor number, mixture ratios, and sample composition. STR profile completeness, allele drop-out rates, and reproducibility across single cells will be evaluated, and a consensus profiling strategy will be applied to improve the reliability of reconstructed genetic profiles.

In parallel, statistical models for the interpretation of single-cell genetic data will be refined, with particular attention to stochastic effects such as allele dropout and amplification variability. Probabilistic frameworks will be adapted to improve the robustness of likelihood ratio calculations, especially in the presence of partial or low-quality profiles.

The proposed approach aims to enhance the resolution and reliability of DNA profiling in complex forensic mixtures and supports the potential integration of single-cell analysis into routine forensic practice.

Authors

• Giorgia Spera (School of Criminal Justice, Faculty of Law, Criminal Justice and Public Administration, University of Lausanne, Lausanne Vaud, Switzerland, Switzerland)
• Franco Taroni (School of Criminal Justice, Faculty of Law, Criminal Justice and Public Administration, University of Lausanne, Lausanne Vaud, Switzerland, Switzerland)
• Diana Hall (Unité de Génétique Forensique, Centre Universitaire Romand de Médecine Légale, Centre Hospitalier Universitaire Vaudois et Université de Lausanne, Ch. de la Vulliette 4, 1000 Lausanne, Switzerland, Switzerland)