Low-copy-number DNA analysis can generate a genetic profile from just a handful of cells, but the results are substantially less reliable than standard DNA testing, and American courts are split on whether to let juries hear them. Standard forensic kits work with about one nanogram of DNA and produce consistent profiles. LCN testing operates below 200 picograms, sometimes from as few as 15 to 30 cells, and the science gets noticeably shakier at those levels.1Croatian Medical Journal. Validity of Low Copy Number Typing and Applications to Forensic Science Whether a judge admits an LCN result usually comes down to two things: what safeguards the specific laboratory built into its protocols, and which admissibility standard the court applies.
What LCN Testing Actually Does
A single human cell contains about six picograms of DNA.2QIAGEN. How Much DNA Does a Human Cell Contain? Standard profiling kits are optimized for around one nanogram of input, roughly 160 cells’ worth. LCN methods push below 100 to 200 picograms. The biological material at that level is invisible, and the lab’s margin for error shrinks accordingly.1Croatian Medical Journal. Validity of Low Copy Number Typing and Applications to Forensic Science
The samples that trigger LCN testing are often called touch or trace DNA. A person grips a doorknob, picks up a tool, or brushes a sleeve, and a few skin cells transfer onto the surface. Investigators swab it hoping to recover something. By the time the swab reaches the lab, there may be almost nothing to work with.
To get a profile from so little material, labs run extra cycles of Polymerase Chain Reaction (PCR) amplification. Standard profiling runs 28 to 30 cycles.3National Institute of Justice. Low Copy Number (LCN) DNA LCN protocols push that to 34 or more. Every added cycle doubles the number of DNA copies, so moving from 30 to 34 cycles is roughly a 16-fold jump in amplification. That extra sensitivity is also what creates the reliability problems the rest of this article walks through.
Why the Results Can Be Wrong
When you amplify a tiny amount of DNA through 34 or more PCR cycles, random chance starts driving the picture that comes out of the machine. Forensic scientists call these random fluctuations stochastic effects. They show up in several ways, and each one can send a case in the wrong direction.
Allele Dropout and Drop-In
Every person carries two copies, called alleles, at each genetic marker. In a clean profile both show up. With LCN samples one allele sometimes fails to amplify, which is called allele dropout, and a true match can look like an exclusion because the profile appears incomplete. The opposite also happens. Allele drop-in means a stray DNA fragment, perhaps left over from a prior sample in the lab or picked up during collection, gets amplified and appears as an extra marker that was never part of the original sample. At sub-200-picogram levels neither artifact is rare.
Contamination the Lab Can Barely See
Standard DNA testing already demands tight contamination controls. LCN testing demands more. Because the method is built to detect a few cells, even a microscopic bit of foreign DNA from a collector’s glove, a shared surface, or a piece of equipment can be amplified alongside the evidence. At 34 cycles the process cannot tell which DNA matters and which drifted in from somewhere else. That makes it genuinely hard to know whether a profile reflects the perpetrator, a first responder, or a technician who worked at an adjacent bench.
Mixture Interpretation
Crime scene samples often contain DNA from more than one person. When the starting material is abundant, analysts can usually separate the contributors by comparing peak heights on an electropherogram. When the starting material is tiny, peaks become uneven, some vanish, and the analyst has to make judgment calls about which markers belong to whom. The more contributors and the less DNA, the more subjective that interpretation becomes.
When an Accurate Profile Still Misleads
Even when the lab correctly identifies whose DNA is on an object, the result can still point at the wrong person. DNA transfers freely between people and surfaces through ordinary contact. Shake someone’s hand and their skin cells travel to the next doorknob you touch. Your DNA can appear on an object you never held, in a room you never entered.
This is not theoretical. In 2012, Lukis Anderson was charged with murder after his DNA was found under a victim’s fingernails. He had an airtight alibi: he was hospitalized for intoxication at the time of the killing. Investigators eventually determined that paramedics who treated Anderson earlier that night later responded to the crime scene and transferred his DNA to the victim.
European authorities spent years hunting the so-called Phantom of Heilbronn, whose DNA appeared at more than 40 crime scenes across Germany, France, and Austria. The phantom turned out to be a factory worker who had unknowingly contaminated the cotton swabs used to collect evidence at each scene. No widespread standards exist to prevent that type of cross-contamination during evidence collection or lab handling, and the extreme sensitivity of LCN testing makes the problem worse. Modern forensic technology can detect deposits from as few as roughly 16 cells, and humans shed tens of thousands of skin cells every day.
How Courts Decide Whether to Admit LCN Results
Before an LCN result reaches a jury, the trial judge must decide whether the underlying science is reliable enough to be admitted. The framework the court applies often decides the outcome.
The Frye Standard
A minority of states still use Frye, which asks a single question: has the technique gained general acceptance within the relevant expert community? If qualified scientists broadly agree the method can produce valid results, the evidence comes in. If the community is still divided, it stays out. How well a particular lab performed the test is not the Frye question.
The Daubert Standard
Most jurisdictions follow Daubert, which gives the trial judge a hands-on gatekeeping role. Instead of polling the scientific community, the judge evaluates the methodology directly by considering whether the technique has been tested, whether it has been subjected to peer review, its known error rate, whether standards exist to control its operation, and whether it has attracted widespread acceptance. Those factors are guidelines, not a rigid checklist, and judges have discretion in how they weigh them.
Federal Rule of Evidence 702
Federal courts and many state courts apply Federal Rule of Evidence 702. The rule was amended in 2023 and now requires the party offering expert testimony to show it is “more likely than not” that the testimony rests on sufficient facts, uses reliable methods, and reflects a sound application of those methods to the case.4Office of the Law Revision Counsel. Federal Rules of Evidence Rule 702 – Testimony by Expert Witnesses That phrasing makes explicit what some courts had been glossing over: the proponent of the evidence carries the burden of proving reliability by a preponderance, not just raising it as a possibility. For LCN DNA, where reliability is the central dispute, the amendment gives judges a clearer basis to exclude results from labs that cannot show their methods actually work at low template levels.
Two Cases That Illustrate the Split
Courts have examined similar underlying science and reached opposite conclusions. Two rulings show how it happens.
People v. Megnath (2010)
A New York state court held that LCN testing performed by the New York City Office of Chief Medical Examiner passed the Frye standard and was admissible.5Justia. People v Megnath The judge gave significant weight to the lab’s use of triplicate testing: each sample ran three separate times, and only markers that appeared consistently across all three runs were reported. That safeguard, the court found, adequately addressed the risks of stochastic effects and contamination inherent in high-sensitivity amplification.
United States v. McCluskey (2013)
A federal court in New Mexico reached the opposite result under Daubert. The court excluded LCN results after finding that the state laboratory had not validated its methods for samples below 250 picograms and could not account for the expected stochastic effects at those levels.6CaseMine. United States v McCluskey The lab admitted its own reliable lower limit was 250 picograms but had tested samples below that threshold anyway. The court held the government failed to carry its burden under Daubert and Rule 702.
What the Split Teaches
The difference was not the science itself but the lab’s preparation. The New York lab in Megnath had years of validation data, published methods, and procedural safeguards. The New Mexico lab in McCluskey tested below its own validated threshold without showing it could handle the consequences. Admissibility fights over LCN DNA are usually fights about what the specific lab did, not about whether LCN testing can ever work in theory.
The Software Behind the Numbers
Traditional DNA mixture interpretation relied on an analyst reading peak heights and making judgment calls. That approach was always somewhat subjective, and it broke down with the kind of low-level, multi-contributor mixtures LCN testing tends to produce. Probabilistic genotyping software was developed to replace those judgment calls with statistical modeling.
The two systems most widely used in American labs are TrueAllele and STRmix.7Federal Judicial Center. Probabilistic Genotyping Systems for Low-Quality and Mixture Forensic Samples Both use Markov Chain Monte Carlo algorithms to simulate thousands of possible genotype combinations and evaluate which combinations best explain the observed data. The software then reports a likelihood ratio: a number expressing how much more probable the observed profile is if the person of interest contributed than if they did not.8National Institute of Justice. Population Genetics and Statistics for Forensic Analysts – Likelihood Ratio
A likelihood ratio of 1 means the evidence is equally consistent with the person being a contributor or not. Above 1 supports inclusion; below 1 supports exclusion. Forensic reports often use verbal scales: 1 to 10 is limited support, 10 to 100 moderate, 100 to 1,000 moderately strong, 1,000 to 10,000 strong, and above 10,000 very strong support for inclusion.8National Institute of Justice. Population Genetics and Statistics for Forensic Analysts – Likelihood Ratio These numbers sound authoritative, but they are only as good as the data and assumptions fed into the software.
The Source Code Problem
TrueAllele and STRmix are proprietary. Their developers treat the source code as a trade secret, which creates a fundamental tension with the adversarial legal system. Defense attorneys argue they cannot meaningfully challenge results produced by software they are not allowed to inspect. Prosecutors and developers answer that peer-reviewed validation studies and external testing adequately demonstrate the software works.7Federal Judicial Center. Probabilistic Genotyping Systems for Low-Quality and Mixture Forensic Samples
Most courts have sided with the prosecution. But the history of New York City’s Forensic Statistical Tool, a probabilistic genotyping program used in roughly 1,350 cases over about five and a half years, is worth remembering. When a federal judge eventually ordered the source code unsealed and independent experts reviewed it, they found coding errors with meaningful impacts on individual case results. The program was discontinued. Software that has never been independently audited can contain errors that peer-reviewed validation studies alone will not catch.
How Defense Attorneys Attack LCN Evidence
Challenges to LCN DNA generally run on three tracks: the lab’s validation, the software’s transparency, and the analyst’s testimony.
Pretrial Admissibility Hearings
The most effective challenges happen before trial. Defense teams should scrutinize the lab’s internal validation studies to see whether the lab actually tested its methods at the template levels present in the case. McCluskey is the blueprint: the lab admitted its reliable threshold was 250 picograms but tested below it anyway, and the court excluded the results.6CaseMine. United States v McCluskey Validation data, proficiency testing records, and quality control documentation are the starting point.
Electronic Discovery
Independent review needs more than the lab report. Defense teams need the raw data files in their native formats: sample files, project files, matrix files, size standard files, and analysis method files. If probabilistic genotyping was used, all timestamped output folders and their contents are necessary to reconstruct the analysis. Without those files, an outside expert cannot verify that the lab’s software produced the reported results.
The Right to Confront the Analyst
The Sixth Amendment’s Confrontation Clause guarantees a criminal defendant the right to cross-examine the witnesses against them. In Melendez-Diaz v. Massachusetts, the Supreme Court held that forensic laboratory reports are testimonial statements, and the analysts who prepared them are witnesses for Sixth Amendment purposes. The prosecution cannot simply submit a lab report; the analyst must be available for cross-examination unless the defendant had a prior opportunity to examine them and the analyst is now unavailable.9Justia. Melendez-Diaz v Massachusetts, 557 US 305 (2009) That matters especially in LCN cases, where much of the analysis involves subjective interpretation. Questioning the analyst about why they included or excluded specific peaks, how they handled ambiguous mixtures, and whether they considered alternatives can expose weaknesses a written report conceals.
The Transfer Defense
Even when the science is solid and the profile is accurate, the defense can argue that the DNA arrived through innocent secondary transfer rather than direct contact with the crime scene. Current forensic technology cannot distinguish between DNA deposited directly by a person and DNA carried there on someone else’s hands. That argument carries weight with juries, especially when the sample quantity is consistent with indirect transfer rather than meaningful physical contact. A prosecutor relying on an LCN result has to explain how the DNA got there, not just whose it is.