Advances in STR chemistry and CE instrumentation, many of them refined in forensic casework, are helping scientists recover usable results from DNA once considered too poor quality and are prompting a reconsideration of when sequencing is really the right tool, explain Nikki Peake and Ellie Kirby.

Across the life sciences, some of the most valuable DNA is often the hardest to work with. Archaeologists analysing centuries-old bone, clinical scientists reading fragmented circulating DNA from blood samples, conservation geneticists typing a museum specimen and forensic analysts working cold cases all struggle with the same underlying problem. The DNA is degraded, present in vanishingly small amounts, or it has come from more than one source.

Ask most lab scientists what to do with a badly degraded, low-level or inhibited DNA sample and the reflex answer is often ‘sequence it’. Next-generation sequencing (NGS) has become the default aspiration for difficult material, while capillary electrophoresis (CE), the workhorse behind short tandem repeat (STR) profiling and fragment sizing, tends to be regarded as legacy technology – adequate for routine samples but insufficient when the DNA is broken, scarce or mixed.

Over the past two years the S TR-CE workflow has been redesigned across chemistry, enzymology and instrumentation, and the sensitivity gap that pushed labs towards sequencing in the first place, has narrowed. Meanwhile the real cost of NGS for targeted questions, in bioinformatics, turnaround and validation burden, has become harder to ignore. The result is labs in forensic science, clinical research and cell science are evaluating their options.

Compromised samples fail for several wellunderstood reasons. Fragmentation, whether from formalin fixation, environmental exposure, age or the naturally short fragments of circulating cell-free DNA (ccfDNA), leaves no intact template for longer STR markers resulting in sub optimal and partial profiles. Low template brings stochastic effects: allele drop-out, imbalance and profiles too weak to call with confidence. Inhibitors carried through from blood, soil or fixatives suppress amplification altogether, and in the case of mixed sample analysis, PCR stutter sits in exactly the positions where a genuine minor allele would appear, obscuring a true result.

Labs have learned to live with those limits on the assumption that STR-CE had hit a ceiling. However, the sensitivity of an STR-CE workflow rests on four independent factors: the number of informative loci, amplicon/marker size, polymerase performance and the dynamic range of the CE instrument. In the last two years all four of these factors have undergone innovative development.

Eight-colour STR chemistry is the most visible development. Two additional dye channels on top of the traditional six create room for up to 35 markers in a single amplification and, more importantly, allow more of them to be designed as mini-STRs.

The sensitivity gap that pushed labs towards sequencing in the first place, has narrowed. Meanwhile the real cost of NGS for targeted questions, in bioinformatics, turnaround and validation burden, has become harder to ignore

The newest eight-colour kits have a mean allele size around 76% smaller than their six-colour predecessors, with most markers under 300 base pairs. That directly translates into more loci recovered from fragmented template. Buffers have also been reformulated to tolerate inhibitor loads that would still stall sequencing library preparation, and built-in quality indicators tell you, within the profile itself, whether a failed result is down to degradation, inhibition or simply no template meaning the first re-work decision is the correct one.

Enzymology has moved in parallel. Engineered reduced-stutter polymerases, such as that in Promega’s Paradyme STR kit, suppress strand slippage to the point where stutter largely disappears below instrument baseline. A peak sitting in a stutter position can then be trusted as genuinely allele, which lowers the analytical threshold making mixture deconvolution simpler, while keeping the simple, quick and cost-effective CE workflow.

Instrumentation supplies the remaining factor. Legacy CE platforms cap at about 32,000 relative fluorescence units, and that ceiling, more than any property of the chemistry, defined how deep a mixture could be read. Current-generation systems, such as the Spectrum CE system, extend dynamic range significantly, some 12 times the legacy ceiling, so the major contributor can tower without saturating while the minor contributor is lifted clear of the baseline.

The combined effect of these advances shows up in the data. Work presented by Eurofins Forensic Services at EAFS and ISHI in 2025 showed an eight-colour kit recovering 9.3 average passing loci from degraded samples at 0.5ng input, against 5.5 for the previous-generation kit, a 47% gain in allele recovery from chemistry alone. Raising input to exploit the wider dynamic r ange pushed that to a 92% improvement over the original baseline. In mixture studies, reliable minor-contributor detection extended from around 5% at standard input to about 1% at high input, territory previously conceded to molecular methods such as NGS.

For forensic laboratories the implications of these advances are clear, more probative information from touch, degraded and inhibited casework, fewer blind repeats and mixture interpretation that is faster and easier to automate, all within accredited workflows. Operational laboratories are reaching the same conclusion.

In mixture studies, reliable minor-contributor detection extended from around 5% at standard input to about 1% at high input, territory previously conceded to molecular methods such as NGS

At the Forensic Genomics Innovation Hub in Southampton, established in 2020 to bridge research and operational casework, scientists have been putting modern CE through its paces. The Hub became the first UK laboratory to install a next-generation multi-capillary Spectrum CE system in late 2023 and have found that on some compromised samples, the latest eight-dye chemistries delivered results equivalent to, or better than, their NGS workflows, in a fraction of the turnaround time and at a fraction of the cost.

The relevance runs well beyond forensics. In clinical research, the clearest example is chimerism monitoring after stem cell transplantation. STR-CE remains the UK standard of care, endorsed by the 2025 British Society of Blood and Marrow Transplantation and Cellular Therapy. However, when DNA from more than one source is present, the limiting factor has been the 1-5% detection floor set by stutter, against 0.1% for qPCR, dPCR, NGS (Yanir et al, 2024), and the inability to improve this due to detector saturation on the CE instrument from the major contributor. However, suppressing stutter at the point of amplification and reading the result on a high-dynamic-range instrument, moves STR towards the sensitivity boundary of these molecular methods without sacrificing sameday turnaround, reproducibility and low cost. Multi-allelic STR markers can also resolve more than two genotypes in a single sample, as in double-cord transplants, which biallelic based methods such as qPCR and dPCR cannot do.

For cell science, biobanking and cell and gene therapy manufacturing, the same capabilities serve a quality control role. Roughly one in five cell lines in circulation is misidentified or cross-contaminated (Hughes et al, 2007; Weiskirchen, 2026). This misidentification is a major contributor to irreproducible clinical research and is estimated to cost $28 billion a year in the United States alone (Freedman et al, 2015). The newly published ISO 23511 standard names STR profiling as the recommended DNA-based identification method for cell lines. A workflow that resolves a 1-2% minor contributor at the speed and cost of a routine STR run, and that names the contaminating line rather than just flagging an error, catches problems concentration checks and sequencing never could.

None of this makes sequencing redundant. Where the question is sequence-level (novel variants, large multi-gene panels, whole genomes) or where a minor component must be tracked below 0.1%, NGS and digital PCR remain the right tools. What has changed is the middle ground. Per-sample NGS costs fall steeply at a larger scale, but only when the run is full: the flow cell costs the same whether it carries eight samples or 96. So labs either batch and let urgent samples queue for days or accept a per-sample cost several times the expected figure; CE does not carry this penalty. Sequencing also moves cost, complexity and time off the bench and into bioinformatics, whereas a CE electropherogram can be read by a scientist or software the same afternoon.

The question is not CE versus NGS, or which technology is more advanced. It is which one fits the sample in front of you and the question you are asking. Sometimes that will be sequencing, but, increasingly often, for damaged and limited DNA, it is worth asking first whether a next-generation STR workflow would answer the question the same day, at a fraction of the cost, on infrastructure and accreditation you already have.

References:

Freedman LP, Cockburn IM, Simcoe TS (2015). The Economics of Reproducibility in Preclinical Research. PLOS Biology 13(6):e1002165.doi:10.1371/journal.pbio.1002165

Hughes P, Marshall D, Reid Y, Parkes H, Gelber C (2007). The costs of using unauthenticated, over-passaged cell lines: how much more data do we need? BioTechniques 43(5):575–586.doi:10.2144/000112598

Moore D, Hack A, Pierce A, Eurofins Forensic Services, presented at EAFS 2025 and ISHI 2025. Promega genetic identity kits and analysis software are For Research Use Only unless otherwise stated.

Weiskirchen R (2026). Genetic Insights into the Economic Toll of Cell Line Misidentification: A Comprehensive Review. Medical Sciences 14(1):25.doi:10.3390/ medsci14010025

Yanir A, et al. (2024). Prospects and Potential for Chimerism Analysis after Allogeneic Hematopoietic Stem Cell Transplantation. Cells 13(11):993. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11172215/

Identity kits and analysis software are For Research Use Only unless otherwise stated.