Completing clinical NGS validation is the step that separates a sequencing assay used for research from one whose results can guide patient care, and it is more involved than the single word "validation" suggests. Analytical validation is not one experiment that an assay passes or fails; it is the systematic demonstration of several distinct performance characteristics, each answering a different question about how the assay behaves, and each requiring its own carefully chosen samples. Understanding validation as this set of parameters, rather than as a single hurdle, is the first step to planning it, and it reveals early the constraint that most often governs the timeline: the availability of well-characterized reference material against which performance can be measured.
This guide covers the analytical validation of a clinical sequencing assay: the performance parameters that must be established and what each demonstrates, the reference materials and truth sets that make measurement possible, the special problem of detecting low-frequency variants, the reportable range and the gaps within it, and the documentation and revalidation triggers that keep a validated assay valid. It addresses analytical validation, whether the assay measures what it claims to measure correctly, which is distinct from clinical validation. The governing guidance throughout is that of the recognized standards bodies, referenced below.
Key Takeaways
|
Validation Parameters and What Each Demonstrates
Analytical validation establishes a defined set of performance characteristics, and naming them precisely matters because each is a specific, testable claim about the assay rather than a general assertion of quality. The recognized guidance, including the joint validation recommendation of the Association for Molecular Pathology and the College of American Pathologists and the corresponding standards from the Clinical and Laboratory Standards Institute, frames these parameters consistently. Each answers a distinct question, and an assay is validated only when all of them have been established for its intended use.
Parameter | What It Demonstrates | Typical Sample Requirement |
Accuracy | That calls match a known correct answer | Reference material with a known truth set |
Precision | That repeat testing gives consistent results, within and between runs | Repeated samples across runs, operators, and days |
Analytical sensitivity (limit of detection) | The lowest variant level reliably detected | Samples with variants at defined, low frequencies |
Analytical specificity | That the assay does not call variants that are not there | Samples with known negative positions; interfering-substance testing |
Reportable range | The regions the assay can reliably report | Coverage analysis across the target, with gap identification |
Table 1. The core analytical validation parameters, each a distinct testable claim with its own sample requirement. The hook made concrete: validation is these parameters together, and several depend on well-characterized reference material.
The pattern the table makes visible is the heart of the hook. Every parameter but the reportable range depends on having samples whose correct answer is already known, and that dependence, not the laboratory work of running the assay, is what usually sets the pace of a validation. Accuracy needs a truth set. Precision needs enough characterized material to test repeatedly. The limit of detection needs samples with variants present at known low frequencies, which are the hardest of all to obtain. Validation planning is therefore, in large part, reference-material planning.
Reference Materials and Truth Sets
A validation can only demonstrate that an assay produces correct answers if there is an independent, trusted source of what the correct answers are, and that source is the reference material and its associated truth set. A truth set is the set of known, high-confidence variant calls for a characterized sample, established by orthogonal methods and community consensus, against which the assay’s output is compared. Without one, there is nothing to measure accuracy against, which is why reference materials are foundational rather than incidental to validation.
The most widely used public reference materials for human sequencing come from the Genome in a Bottle consortium, coordinated by the National Institute of Standards and Technology, which has produced extensively characterized human genomes, along with the National Institute of Standards and Technology’s own reference materials derived from them. These provide high-confidence truth sets across much of the genome and are the common foundation for accuracy validation. They are supplemented, where needed, by well-characterized cell lines and by materials engineered to carry specific variants of clinical interest. The practical difficulty is that these public materials, however excellent, do not cover every variant a clinical assay must detect, and the gap between what is characterized and what a given assay needs to validate is exactly where reference material becomes the binding constraint.
The Truth Set Is the Bottleneck The limiting resource in most clinical sequencing validations is not instrument time or staff effort but the availability of reference material with a trusted truth set for the specific variants the assay must detect. Public consortium materials cover common, well-studied regions well, but a clinical assay often has to detect rare variants, difficult genomic regions, or specific alterations for which no characterized reference sample exists. Where none exists, the laboratory must create validation material, by engineering, by mixing, or by exhaustively characterizing a sample with orthogonal methods, and that work, not the assay run, is what dominates the validation timeline. Plan the reference-material sourcing first, because it is the critical path. |
Lab Quality Management Certificate
The Lab Quality Management certificate is more than training—it’s a professional advantage.
Gain critical skills and IACET-approved CEUs that make a measurable difference.
Limit of Detection for Low-Frequency Variants
The limit of detection is the parameter that most often determines whether an assay is fit for its clinical purpose, and it is the hardest to establish, because it requires demonstrating reliable detection at low variant frequencies where the signal approaches the noise. For many clinical applications, particularly the detection of variants present in only a small fraction of the molecules in a sample, such as a mutation in a minority of tumor cells or circulating tumor material, the clinically important variants are exactly the low-frequency ones, so the limit of detection is not a technicality but the parameter on which the assay’s usefulness turns.
Establishing it requires samples in which the variant of interest is present at known, low frequencies, typically created by mixing characterized materials in defined proportions to produce a dilution series spanning the range around the intended detection threshold. The assay is then challenged to detect the variant across that series, establishing the lowest frequency at which detection is reliable. This is demanding on both sides: it requires the depth of sequencing sufficient to distinguish true low-frequency variants from background error, a relationship between coverage and detection developed in the coverage and cost guidance, and it requires characterized materials at the low frequencies themselves, which returns once more to the reference-material constraint. An assay’s claimed limit of detection is a specific, validated number, and it defines the floor below which the assay must not report, which is why establishing it rigorously is central to clinical fitness.
Interested in lab leadership?
Reportable Range and Gap Analysis
An assay does not perform uniformly across its entire target, and the reportable range is the honest definition of where it performs well enough to report a result, which requires identifying the places where it does not. Sequencing coverage and performance vary across the genome: some regions sequence cleanly and are reliably callable, while others, because of high similarity to other regions, extreme base composition, or repetitive structure, are covered poorly or ambiguously and cannot support a reliable call. The reportable range is the defined set of regions the assay can reliably report, and everything outside it must be excluded from reporting or explicitly flagged.
Gap analysis is the process of finding these under-performing regions within the assay’s nominal target, and it is essential because an undetected gap is a place where the assay can silently fail to detect a variant, reporting a region as negative when in truth it was never adequately examined. A validated assay has a documented reportable range and a known, managed set of gaps, so that a clinician receiving a result knows what was and was not reliably interrogated. Regions that fall in a gap but are clinically important may need to be filled by an orthogonal method, and defining how gaps are identified, managed, and communicated is a required part of validation, not an optional refinement.
Documentation and Revalidation Triggers
A validation exists only insofar as it is documented, because in a regulated setting an undocumented validation is indistinguishable from no validation at all, and the documentation is what an inspector, an accreditor, and the laboratory itself rely on to trust the assay’s results. The validation record captures each parameter, the samples and reference materials used, the acceptance criteria, the results, and the conclusion that the assay met its intended use, assembled into a validation summary that is retained and available. This documentation is produced against the recognized guidance and is what accreditation to standards such as those of the College of American Pathologists is assessed against, as part of the broader compliance framework covered in Quality and Compliance in NGS Labs: From Research Use to Regulated Testing.
A validated assay is validated only for what it was validated on, which is why defined changes trigger revalidation. A change to the assay chemistry, the instrument or platform, the bioinformatics pipeline, the reference genome, or the intended specimen types can alter performance, and any such change requires assessing, and usually re-demonstrating, the affected parameters, because the original validation no longer necessarily holds. This is the same principle that governs instrument qualification, where the qualified state must be maintained and re-established after significant change, described in Installation and Instrument Qualification: IQ, OQ, and PQ for Sequencers. Defining in advance which changes trigger revalidation, and to what extent, is part of a mature validation program, and it connects directly to the ongoing performance monitoring, proficiency testing, and quality control that keep a validated assay trustworthy over time, developed in Proficiency Testing and External Quality Assessment for Sequencing Labs. Accreditation requirements that govern the validated assay in operation are covered in CLIA, CAP, and ISO 15189: NGS Lab Accreditation Requirements. How validation fits the whole quality and operational picture is in Next-Generation Sequencing in the Lab: A Manager’s Guide to Building, Budgeting, and Scaling NGS Capacity.
References and Guidance
The standards bodies and reference-material frameworks referenced in this article. Standards documents are available from their issuing organizations; the primary literature citations are given for the peer-reviewed guidance.
Jennings LJ, Arcila ME, Corless C, et al. “Guidelines for Validation of Next-Generation Sequencing–Based Oncology Panels: A Joint Consensus Recommendation of the Association for Molecular Pathology and College of American Pathologists.” The Journal of Molecular Diagnostics, 2017;19(3):341–365.
Clinical and Laboratory Standards Institute. Human Genetic and Genomic Testing Using Traditional and High-Throughput Nucleic Acid Sequencing Methods. 3rd ed. CLSI guideline MM09. Wayne, PA: CLSI; 2023. (Replaces the earlier MM09-A2, 2014.)
Aziz N, Zhao Q, Bry L, et al. “College of American Pathologists’ Laboratory Standards for Next-Generation Sequencing Clinical Tests.” Archives of Pathology & Laboratory Medicine, 2015;139(4):481–493.
Roy S, Coldren C, Karunamurthy A, et al. “Standards and Guidelines for Validating Next-Generation Sequencing Bioinformatics Pipelines: A Joint Recommendation of the Association for Molecular Pathology and the College of American Pathologists.” The Journal of Molecular Diagnostics, 2018;20(1):4–27.
Zook JM, Catoe D, McDaniel J, et al. “Extensive sequencing of seven human genomes to characterize benchmark reference materials.” Scientific Data, 2016;3:160025. Genome in a Bottle (GIAB) Consortium and reference materials, National Institute of Standards and Technology (NIST).
This article was produced under Lab Manager's AI Editorial Guidelines.
