The research team, led by Zirui Lu, Changle Li, Jinghui Yang, Xiaotong Wang, Jing Jin, and Fei Huang, systematically exposed dental samples to temperatures ranging from 200 to 500 degrees Celsius for periods of 10 and 20 minutes, then attempted to recover three different classes of genetic markers. Short tandem repeats, or STRs, are the workhorses of forensic identification, generating the familiar DNA profiles used in criminal databases. Single nucleotide polymorphisms, or SNPs, offer an alternative that can succeed even when DNA is badly fragmented because the target regions are shorter. Mitochondrial DNA, or mtDNA, exists in hundreds to thousands of copies per cell and has traditionally been the fallback for degraded samples. By tracking all three simultaneously, the researchers could construct a thermal survival hierarchy that had never been quantified this precisely.

The results reveal a startling sensitivity to time. At 200 degrees Celsius, extending the heating duration from 10 to 20 minutes caused STR detection to plummet from 97.09 percent to 81.39 percent, SNP detection to drop from 99.53 percent to 91.08 percent, and mtDNA sequencing coverage to collapse from an almost perfect 99.96 percent to a dismal 27.26 percent. In other words, simply doubling the exposure time — even at a temperature well below that of a typical house fire — was enough to wipe out nearly three-quarters of the recoverable mitochondrial genome. Within the 200 to 220 degree Celsius range, the duration of exposure exerted a stronger effect on DNA integrity than the temperature itself, a finding that challenges the intuition that heat intensity should be the dominant variable.

The critical cliff edge came at 250 degrees Celsius. After just 10 minutes at that temperature, no genetic markers of any type could be recovered from the dental pulp. By contrast, at 220 degrees Celsius for 10 minutes, nuclear DNA retained full detectability at 100 percent and mtDNA coverage still reached 76.48 percent. Prolonging the treatment to 20 minutes at 220 degrees drove STR detection down to 66.99 percent, SNP detection to 99.3 percent, and mtDNA coverage to 28.84 percent. For fire investigators, this defines a practical threshold: teeth that have been heated beyond roughly 250 degrees for any meaningful duration are unlikely to yield a conventional DNA profile, and sampling strategies must be planned accordingly.

Perhaps the most counterintuitive finding concerns mitochondrial DNA. Forensic practice has long treated mtDNA as the most durable target, relying on its enormous copy number to survive conditions that destroy nuclear DNA. This study demonstrates the opposite under thermal stress: mtDNA is actually more heat-vulnerable than nuclear DNA. The authors attribute this to three structural weaknesses. Unlike nuclear DNA, which is wrapped around histone proteins that shield it from damage, mtDNA lacks histone protection entirely. Its genome is also rich in adenine and thymine, the two nucleotides joined by only two hydrogen bonds rather than three, making those regions easier to denature. Finally, heat disrupts the circular conformation of the mitochondrial genome, breaking the closed loop that normally protects it from exonuclease attack. The high copy number, it turns out, simply cannot compensate for these structural drawbacks.

The study also compared the protective value of the tooth’s internal environment against its exterior. Bloodstains deposited on tooth surfaces — representing the kind of trace evidence that might cling to recovered remains — proved far less thermally tolerant than the pulp sealed inside. After 20 minutes at 200 degrees Celsius, STR and SNP detection rates in the surface bloodstains fell to 72.58 percent and 88.38 percent respectively, and mitochondrial DNA became completely undetectable. At 220 degrees for 10 minutes, the bloodstains still yielded an STR detection rate of 97.5 percent and full SNP detection at 100 percent, but mtDNA coverage dropped to a mere 16.9 percent. After 20 minutes at that same temperature, every marker type was unrecoverable. The mineralized shell of the tooth, in short, shields the pulp far more effectively than any surface stain can protect itself, reinforcing the long-standing recommendation that dental pulp should be the sampling site of choice.

Technically, the team employed two complementary genotyping strategies: capillary electrophoresis of STR markers, the gold-standard method used in forensic laboratories worldwide, and massively parallel sequencing, or MPS, which reads millions of DNA fragments simultaneously and can target short SNP amplicons suitable for degraded samples. They calculated marker detection rates across the temperature and time matrix and quantified mtDNA sequencing coverage to assess how much of the mitochondrial genome remained readable. The short-fragment MPS-SNP approach proved the most resilient across the thermal gradient, and the authors recommend it as the preferred strategy for severely degraded specimens where conventional STR profiling would fail.

The implications reach back to some of the most challenging identification efforts in modern history. DNA profiling of burned and fragmented remains has been central to victim identification after the World Trade Center attacks, the Bali and Jakarta bombings, and countless aviation disasters. Each of these efforts confronted the same fundamental problem: knowing which tissue to sample and which genetic technology to deploy when heat has already done its damage. By establishing precise temperature-time thresholds and a clear stability hierarchy — SNP greater than STR greater than mtDNA — the new study gives forensic laboratories a decision framework grounded in controlled experimental data rather than anecdotal experience. Exposure duration, the authors argue, should be treated as a formal parameter when assessing specimen quality, alongside the gross thermal appearance of the tooth itself.

The research was approved by the Scientific Research Ethics Committee of the Forensic Appraisal Center of the Ministry of Public Security and was funded through several Chinese national and institutional programs, including the National Key Research and Development Program of China. Its practical guidance is straightforward: prioritize dental pulp for DNA sampling from fire-exposed dental specimens, factor in the likely duration of heating when evaluating what remains recoverable, and turn to short-fragment MPS-SNP genotyping when degradation is severe. As wildfires intensify and urban fires remain a persistent threat worldwide, the humble tooth — and the delicate balance of time and temperature written into its DNA — may determine whether families ever receive answers about their lost loved ones.

Subject of Research: Forensic DNA recovery of STR, SNP, and mitochondrial DNA from teeth after thermal exposure

Article Title: Comparative study of STR, SNP, and mtDNA recovery in teeth after thermal exposure

Article References: Lu, Z., Li, C., Yang, J., Wang, X., Jin, J., & Huang, F. (2026). Comparative study of STR, SNP, and mtDNA recovery in teeth after thermal exposure. International Journal of Legal Medicine. https://doi.org/10.1007/s00414-026-04005-7

Image Credits: AI Generated

DOI: 10.1007/s00414-026-04005-7

Keywords: forensic science, DNA profiling, teeth, thermal degradation, STR, SNP, mitochondrial DNA, dental pulp, massively parallel sequencing, fire victims, DNA identification, International Journal of Legal Medicine

The research team, led by Maryam Sharafi Farzad and colleagues at the University of Copenhagen, set out to validate the Precision ID mtDNA Whole Genome Panel from Thermo Fisher Scientific specifically for telogen hair analysis and to implement it in routine forensic casework. Mitochondrial DNA offers a crucial advantage in this context: cells contain many more copies of mitochondrial DNA than nuclear DNA, and the molecule is far more resistant to degradation. Although mtDNA is maternally inherited and therefore lacks the discriminatory power to identify a single individual, it remains highly valuable for maternal lineage analysis, missing-person investigations, disaster victim identification, and cases where only degraded biological material is available. Traditionally, mtDNA typing relied on Sanger sequencing of the control region alone, a labor-intensive and costly approach with limited discriminatory power. Massively parallel sequencing technologies such as the Precision ID panel now allow entire mitochondrial genomes to be sequenced quickly and affordably, dramatically increasing both sensitivity and information content.

To establish the sensitivity of the workflow, the researchers prepared dilution series from NIST Standard Reference Material DNA and six reference samples, spanning an extraordinary range from 3 to 3,600 mitochondrial DNA copies per microliter. The results were striking. Complete mitochondrial genome sequences were consistently recovered in 99 percent of samples across the dilution series at DNA input levels as low as 47 mtDNA copies, demonstrating remarkable robustness under low-template conditions. Full profiles were obtained in half of the samples containing only 23 copies, and even some samples at 12 copies produced partial results. Statistical modeling using log-log linear regression revealed that library yield scaled nearly proportionally with mtDNA input in the NIST series, with a doubling of input DNA producing an approximately 2.16-fold increase in library output. Read depth increased linearly up to roughly 560 to 1,235 copies per reaction before plateauing due to saturation, underscoring the importance of accurate library quantification and normalization.

The heart of the study involved 100 volunteers who each provided 20 shed hairs collected from their own clothing, with 98 of them also contributing buccal swab samples for comparison. Microscopic examination identified the club-shaped roots characteristic of telogen hairs, and one hair per donor was selected for analysis. Quantification revealed that 98 of the 100 hair extracts met or exceeded the manufacturer-recommended input of 3,000 mtDNA copies, and both small and large mtDNA quantification targets were detected in every single sample. In contrast, 25 hairs contained no detectable nuclear DNA at all, and most of the remainder showed concentrations below 6.8 picograms per microliter with extensive degradation. The researchers also found that reducing the post-extraction volume from 50 to 30 microliters significantly increased both mtDNA copy number and sequencing read depth, a simple procedural change that boosted sensitivity without introducing contamination. Notably, donor variables such as age, gender, hair dye use, and washing habits showed no statistically significant association with mtDNA quantity or quality.

Comparing the mitochondrial genomes sequenced from 91 paired hair and buccal samples revealed fascinating subtleties of mitochondrial biology. Initial sequence concordance between hairs and their corresponding buccal references was 93.4 percent. However, when the team applied the interpretation guidelines of the International Society for Forensic Genetics and excluded low-confidence discrepancies associated with high strand bias, heteroplasmic variants, and sequencing artifacts, concordance rose to a perfect 100 percent. This finding is critically important for forensic practice: apparent mismatches between a hair and a reference sample frequently reflect technical limitations and tissue-specific heteroplasmy rather than true genetic differences. Heteroplasmy, the coexistence of multiple mitochondrial variants within an individual, arises from bottleneck effects and drift during tissue development, and previous research has shown that it can drift notably in hair. According to ISFG guidelines, a single mismatch or heteroplasmic difference should never be interpreted as evidence of exclusion, and this study provides empirical weight for that principle in the context of whole mitochondrial genome sequencing.

Reproducibility testing further strengthened the case for implementation. The researchers amplified and sequenced duplicate libraries from 43 telogen hair samples and 46 buccal samples, finding over 99 percent concordance of the software-classified true variants between replicates. The handful of discrepancies involved shifts between point and length heteroplasmy classifications, insertion-length differences at homopolymeric C-tracts, and low-frequency artifacts at known hotspots, none of which affected haplotype interpretation. Haplogroup assignments remained identical across all replicates and dilution levels, confirming that reduced template input did not compromise phylogenetic classification. The data analysis itself was also highly repeatable, with two independent runs of the Converge analysis software producing identical results across every measured parameter, from variant frequencies and read depths to strand bias metrics and variant state classifications.

Contamination control, a perennial concern in low-copy-number DNA work, was rigorously assessed. Of 16 negative controls processed during the study, none produced a complete mitochondrial DNA sequence. Six controls showed low-level mtDNA signal consisting of only two to eight variants, predominantly common polymorphisms, and none of these partial profiles corresponded to any of the analyzed samples, the NIST control, or laboratory personnel. Several variants in the controls were traced to primer binding regions, indicating primer variants rather than true biological signal. All work was conducted in a dedicated laboratory following ISFG and SWGDAM guidelines under the ISO 17025 validation framework, and every extraction, quantification, and sequencing batch included negative controls, providing a template for laboratories considering adoption of the method.

Perhaps most compelling for investigators is the performance of the validated workflow on real evidence. The team sequenced 11 archived DNA extracts from hairs collected in 2010 from concluded criminal cases, samples that had previously been deemed unsuitable for STR profiling because they contained no detectable or less than 7 picograms per microliter of nuclear DNA. Remarkably, these aged forensic samples yielded complete mitochondrial genome sequences with an average read depth of 1,037-fold, an exceptionally high coverage figure. Across the three fully reported casework samples, 102 variants were identified, 98 of which were confirmed as true variants with high confidence scores, and no nuclear mitochondrial DNA segments, or NUMTs, were detected to confound the analysis. The success demonstrates that even DNA extracts stored for over a decade can return informative whole mitochondrial genomes, offering hope for cold case reviews.

The study is not without caveats that laboratories must respect. A few amplicons, particularly those covering mitochondrial regions 8622 to 8796, 1317 to 1491, and 1948 to 2111, consistently underperformed across sample types, mirroring poor performance reported in earlier evaluations of the panel. Length heteroplasmy artifacts at homopolymeric stretches such as positions 309 and 11038 appeared with variant strand bias exceeding 0.7, requiring careful filtering. At very low template levels, below roughly 47 copies, stochastic effects increased the frequency of unlikely and unclear variant calls. The authors emphasize that reliable results from low-copy samples demand increased read allocation, replicate sequencing, manufacturer-recommended run thresholds, and stringent filtering of negative controls and NUMTs. The laboratory will continue using an input of 3,000 mtDNA copies for casework despite evidence that lower inputs can suffice, retaining a conservative margin for evidentiary work.

Taken together, the findings position whole mitochondrial genome sequencing of telogen hairs as a validated, sensitive, and contamination-controlled method ready for routine forensic implementation. With more than 85 percent of hair evidence in some laboratories currently unanalyzable by STR profiling, the ability to recover complete mitochondrial genomes from shed hairs with high read depth and near-perfect concordance when properly interpreted represents a meaningful expansion of the forensic toolkit. As massively parallel sequencing continues its migration from research laboratories into accredited casework units, hairs that once sat silently in evidence storage may soon speak, narrowing gaps in investigations, strengthening maternal lineage identifications, and giving cold cases a second chance at resolution.

Subject of Research: Validation of whole mitochondrial genome sequencing of shed telogen hairs for forensic casework

Article Title: Precision ID mtDNA Whole Genome Panel and sequencing of telogen hairs – perspectives for validation and implementation in casework

Article References: Sharafi Farzad, M., Claessens, F., Petersen, C. B., Pereira, V., Morling, N., & Mogensen, H. S. (2026). Precision ID mtDNA Whole Genome Panel and sequencing of telogen hairs – perspectives for validation and implementation in casework. International Journal of Legal Medicine. https://doi.org/10.1007/s00414-026-03992-x

Image Credits: AI Generated

DOI: 10.1007/s00414-026-03992-x

Keywords: mitochondrial DNA, telogen hairs, massively parallel sequencing, Precision ID mtDNA Whole Genome Panel, forensic casework, mtDNA heteroplasmy, low-template DNA, STR profiling, profile concordance, contamination control, ISFG guidelines, hair shaft evidence