Touch DNA and Trace Biological Material

Human skin sheds approximately 30,000 to 40,000 corneocytes per hour from the total body surface. The outermost cells of the stratum corneum are anucleate: they have lost their nuclei as part of terminal differentiation, which means they contribute no DNA. The cells that matter forensically are the nucleated cells from slightly deeper layers, the granular and spinous layers of the epidermis, that are also displaced and deposited during contact. These cells retain nuclei and carry the full nuclear genome.

The number of nucleated cells deposited in a single contact varies enormously between individuals. Some people are consistently high shedders, depositing hundreds to thousands of cells per touch; others are low shedders, depositing so few cells that a detectable profile is rarely achieved. This inter-individual variation is partly genetic, partly influenced by skin hydration, skin condition (cuts or abrasions increase shedding), recent hand washing, and the time elapsed since the last contact with a surface. A person who has just washed their hands may shed fewer cells on the next surface they touch.

The mechanics of contact also matter. Gripping an object with sustained pressure transfers more cells than a brief brush. Twisting or sliding motion across a surface transfers more cells than static contact. Objects that are manipulated repeatedly, such as a steering wheel or firearm grip, accumulate cells from multiple contact events and from multiple individuals, creating complex mixtures.

Primary transfer is the direct deposition of skin cells from a person's surface onto an object during contact. The amount transferred depends on the factors described above. Once deposited, those cells begin to degrade through physical, chemical, and biological mechanisms. UV radiation damages DNA by creating pyrimidine dimers. Humidity promotes hydrolysis and microbial colonisation. Heat accelerates both. Oxidising agents in the environment degrade the phosphodiester backbone. On outdoor surfaces in warm climates, a touch DNA sample may become undetectable within hours in direct sunlight, while the same cells on an indoor surface protected from light and moisture may persist for months.

Substrate type is one of the strongest predictors of persistence. On glass and stainless steel, cells adhere relatively poorly but are protected from moisture absorption and microbial penetration. On porous substrates like wood, cotton, and paper, cells are retained more firmly but are exposed to moisture ingress and microbial activity within the material. Plastic varies: smooth dense polymers behave similarly to glass; textured or open-weave polymers behave more like fabric. When evidence reaches the laboratory, the packaging history matters as much as the substrate: paper bags allow moisture exchange and slow degradation; airtight plastic bags trap humidity and accelerate microbial growth.

Persistence studies have shown that on indoor hard surfaces, touch DNA profiles can be recovered after four weeks or longer. Outdoor substrates in temperate climates typically yield profiles for one to seven days, depending on weather. These are averages and the variance is high: in a documented series of outdoor experiments in Australia, some samples on metal poles failed at 24 hours while others on the same pole type yielded profiles at 14 days.

Secondary transfer occurs when cells that were deposited on one surface (the primary surface) are subsequently moved to another surface (the secondary surface) via an intermediate contact, without the original contributor ever touching the secondary surface. Classic scenarios include: a person shakes hands with an individual who then handles a weapon; a garment worn by a contributor is worn by another person; or a vehicle seat retains cells from one occupant that transfer to clothing of another.

Laboratory studies have demonstrated secondary transfer across a range of substrate combinations. A handshake with sustained grip can transfer enough cells from Person A through Person B's hands to produce an STR profile of Person A on an object touched only by Person B, especially when Person B touches the object immediately after the handshake. Fabric-to-fabric transfer during shared clothing use produces similar results. Tertiary transfer, a third step in the chain, has also been documented, though the yield per step drops substantially at each stage.

The quantity of DNA transferred at each stage decreases, but this does not make secondary transfer negligible in casework. High-shedder individuals produce primary deposits that are large enough to survive the efficiency losses of secondary transfer and still yield a detectable profile. The forensic scientist must consider: when did contact with the primary surface occur, how much time elapsed before the secondary contact, how many intermediate surfaces were involved, and what is the estimated cell-transfer efficiency for each substrate combination. These factors feed into a probability-based evaluation rather than a binary presence-or-absence conclusion.

The choice of collection method significantly affects the number of cells recovered and the quality of the resulting profile. The double-swab technique is the most widely adopted method for smooth non-porous surfaces such as glass, metal, and smooth plastic. A swab dampened with sterile water is rubbed firmly over the target area for 30 seconds, then a dry swab is immediately passed over the same area. The wet swab rehydrates and loosens cells that have dried onto the surface; the dry swab absorbs the moisture and captures the suspended cells. Both swabs are packaged together and submitted for extraction.

For porous substrates such as clothing, upholstery, and rope, cutting a section of the material is preferred over swabbing, because swabbing leaves a proportion of cells embedded in the fibres. The cutting should include all layers of the fabric at the area of interest. For large items where sampling must be targeted, the location with the highest expected contact (the inner cuff of a glove, the collar of a shirt, the grip area of a ligature) is sampled first. Multiple samples from the same item are packaged separately to preserve spatial information about cell distribution.

Swabs must be air-dried before packaging to prevent cell degradation from trapped moisture. Refrigeration at 4 degrees Celsius or freezing at minus 20 degrees Celsius is recommended for long-term storage before analysis. Contact with the evidence item by scenes of crime officers or laboratory staff without appropriate PPE can deposit examiner DNA, generating a contamination profile. Touch DNA scenes require full glove-plus-mask PPE from the first responder onward, and all personnel involved in item handling must be elimination-profiled.

Touch samples routinely contain less than 100 picograms of DNA, well below the 1 to 2 nanogram threshold typically required for standard 28-cycle STR amplification to produce a reliable profile. Below this threshold, PCR amplification becomes stochastic: the probability that a given allele is represented in the extracted template decreases, and whether it amplifies or not is partly a matter of chance. This produces the three canonical stochastic effects: allelic dropout (one allele at a heterozygous locus fails to amplify), allelic drop-in (a spurious allele appears, typically from trace contamination amplified to detectable levels), and peak height imbalance (the two alleles at a locus produce peaks of very different heights, mimicking a heterozygous-to-homozygous ratio).

Sensitivity can be increased by extending the number of PCR cycles from 28 to 34, the approach originally called low copy number (LCN) DNA by the UK Forensic Science Service. LCN increases the probability of amplifying low-abundance alleles but proportionally amplifies trace contaminants and drops-in. International practice now recommends replicate amplification, where the same extract is amplified independently two or three times, and only alleles that appear consistently across replicates are included in the consensus profile. This approach reduces false-positive allele calls at the cost of increased analytical time and consumable use.

Many touch surfaces carry cells from more than one person, producing a mixed DNA profile. Mixture interpretation from low-template touch DNA is among the most computationally demanding tasks in forensic biology. Probabilistic genotyping software packages, such as STRmix (New Zealand and Australia), TrueAllele (US), and Euroformix (Europe), calculate likelihood ratios under competing hypotheses about who contributed to the mixture and in what proportions. These systems are now accepted in courts across multiple jurisdictions, including under Daubert standards in federal US courts and under the Criminal Practice Directions in England and Wales. The Bharatiya Sakshya Adhiniyam 2023 (which replaced the Indian Evidence Act 1872) provides the framework for expert opinion admissibility in Indian courts, though specific probabilistic genotyping guidelines remain under development by agencies such as the Central Forensic Science Laboratory network.

A touch DNA profile, even a complete one matching a person of interest, does not by itself establish that the person touched the item at the time of the offence. The forensic scientist must consider four questions: Is the profile genuine, that is, not a contamination or a laboratory artefact? If genuine, did the contributor deposit cells directly or via secondary transfer? When was the deposit made, relative to the time of the alleged offence? And how many contributors are represented in the profile? Only after addressing all four questions can a meaningful statement about evidential weight be made.

Likelihood ratio (LR) frameworks are the standard tool for reporting DNA match weight in most forensic science systems. The LR compares the probability of the observed DNA evidence if the suspect is a contributor against the probability of the same evidence if an unknown unrelated person is the contributor. For touch DNA mixtures, the LR is calculated by probabilistic genotyping software over all combinations of possible genotypes. An LR of one million means the evidence is one million times more probable under the prosecution hypothesis than under the defence hypothesis. Courts in the UK, Australia, the Netherlands, and the US now routinely receive LR-based DNA evidence; Indian courts are moving toward similar frameworks under the Digital Personal Data Protection Act 2023 and accompanying forensic science standards discussions.

The forensic scientist's statement should explicitly address the activity level, not only the source level. Source-level evidence states only that the contributor's DNA was detected on the item. Activity-level evidence evaluates whether that finding is more consistent with the contributor having direct contact with the item (prosecution hypothesis) or with an innocent explanation such as secondary transfer (defence hypothesis). Activity-level reporting is endorsed by the ENFSI DNA Working Group in Europe and by the UK Forensic Science Regulator's codes of practice. This approach is gradually being adopted in jurisdictions including Canada, Australia, and New Zealand.