RESUMEN
Tapelifting is a common strategy to recover touch DNA deposits from porous exhibits in forensic DNA casework. However, it is known that only about 30 % of tapelifts submitted for DNA analysis in operational forensic laboratories yield profiles suitable for comparison or upload to a searchable database. A reliable means to identify and remove non-probative tapelifts from the workflow would reduce sample backlogs and provide significant cost savings. We investigated whether the amount of macroscopic or microscopic fluorescence on a tapelift following staining with Diamond Nucleic Acid Dye (DD), determined using a Polilight and Dino Lite microscope respectively, could predict the DNA yield and/or the DNA profiling outcome using controlled (saliva), semi-controlled (finger mark) and uncontrolled (clothing) samples. Both macroscopic and microscopic DD fluorescence could predict DNA yield and profiling outcome for all sample types, however the predictive power deteriorated as the samples became less controlled. For tapelifts of clothing, which are operationally relevant, Polilight fluorescence scores were significantly impacted by clothing fibres and other non-cellular debris and could not be used to identify non-probative samples. The presence of less than 500 cells on a clothing tapelift using microscopic counting of stained corneocytes was identified as a potential threshold for a non-probative DNA profiling outcome. A broader examination of the reliability of this threshold using a casework trial is recommended. Due to the labour intensiveness of microscopic cell counting, and the increased risk of inadvertent contamination, automation of this process using image software in conjunction with artificial neural networks (ANN) should be explored.