Q30: What Metrics Really Matter in Clinical Oncology NGS Testing?

The Q30 score gained traction for legitimate reasons. When next-generation sequencing began entering clinical use, whole-genome and whole-exome sequencing were among the prominent applications, alongside targeted sequencing approaches, particularly in oncology, Procurement teams learned to ask for it, not because it is the right number but because it was an easy number to ask for, tenders started requiring it, and the habit solidified before the more relevant sensitivity and specificity metrics for targeted oncology panels became the norm.

For NGS instrument vendors whose platforms perform well on Q30, it is a flattering number and a familiar one to customers. This metric may offer a single comparable figure across instrument vendor proposals, although direct comparability can be limited because quality-score calibration and error profiles may differ across platforms.. Nobody set out to use a misleading metric; the metric just never got updated when the clinical use case shifted from broad genomic surveying to targeted variant detection.

The problem comes when Q30 is treated as a universal quality requirement across sequencing technologies, clinical contexts, and assay designs. Oncology NGS is not a raw-read generation exercise; it is a task of reliably detecting and classifying genetic alterations at the required levels of detection. What matters at the end of the workflow is not an abstract base-quality profile, but a reliable variant call.

That distinction is especially important when comparing different sequencing technologies, methods, application, and chemistries. 

For oncology NGS evaluations, the more relevant comparison is therefore the performance of the complete workflow in detecting the variants it is intended to detect, rather than a single raw-read quality metric considered in isolation.

The most clinically meaningful quality question in oncology NGS is direct: Can the workflow detect the variant when it is present, and avoid detecting it when it is absent or below the relevant level of detection?

That question is captured through validation metrics such as analytical sensitivity, analytical specificity, positive and negative percent agreement, precision, reproducibility, positive predictive value, and limit of detection. These metrics evaluate performance at the level where clinical and laboratory decisions are made.

This is also why evaluation of NGS tests should address the intended variant classes. A workflow may perform well for single nucleotide variants but require separate evidence for insertions and deletions, copy number alterations, fusions, splice variants, or complex biomarkers. AMP/CAP recommendations for cell-free DNA assays similarly emphasize analytical validation, including sensitivity, specificity, precision, reproducibility, and detection limits suitable for low-frequency variants in plasma and other cfDNA contexts [1].

That framework is more informative than Q30 alone because oncology specimens are often difficult. FFPE-derived nucleic acids may be fragmented or chemically modified. Small biopsies may provide limited input. Cytology samples may vary in tumor content. Plasma samples may contain low tumor fractions. In those settings, a single base-quality metric cannot answer whether the workflow is clinically reliable. Targeted amplicon-based NGS panels can be useful for workflows involving challenging sample types because some designs can accommodate limited nucleic acid input and can be optimized for coverage uniformity.

Whole-genome sequencing and broad research sequencing are often judged heavily on global read characteristics because they aim to survey large genomic regions. In those settings, raw read quality can be a central performance measure.

Targeted oncology NGS has a different purpose. It concentrates sequencing capacity on selected genomic regions so that clinically relevant or biologically relevant targets can be sequenced at high depth. That depth can support detection of low-frequency variants, provided the workflow has appropriate sample preparation, chemistry, controls, bioinformatics, and validation.

This is why raw read count or Q30 percentage can be incomplete or even distracting when used alone. A workflow with strong raw-read metrics is not automatically a better oncology assay. If it lacks validated detection across relevant variant types, has uneven coverage, requires more tissue than is typically available, struggles with degraded FFPE samples, or does not meet required turnaround times, the raw metric does not solve the clinical problem.

Conversely, an oncology workflow should be viewed favorably when it demonstrates accurate and reproducible detection at defined thresholds using the sample types and variant classes relevant to its intended use. For clinical customers, this is the decisive evidence.

For laboratories, procurement teams, and tender evaluators, one approach is to distinguish raw sequencing-output metrics from clinical-performance metrics.

Raw sequencing metrics can be useful for troubleshooting and quality monitoring. They can also help laboratories understand whether a run behaved as expected. But raw metrics become less useful when they are pulled out of context and used as universal purchasing criteria.

Clinical oncology NGS is a workflow. Specimen acquisition, fixation, extraction, library preparation, target enrichment, sequencing chemistry, depth, uniformity, controls, bioinformatics, annotation, and reporting all contribute to the final result. A weakness at any stage can affect performance. A strength at one stage, like raw read sequencing accuracy, cannot compensate for a lack of validation at another.

That is why clinical performance should be assessed at the level of the reported result. The most relevant evidence asks whether the workflow produces accurate calls for the variants it claims to detect. It asks whether the workflow performs consistently across sites and operators. It asks whether low-frequency variants can be detected at the stated LoD. It asks whether negative samples remain negative. It asks whether fusions, CNVs, indels, and other variant classes are validated appropriately.

This is also a more direct way to compare technologies. Different sequencing chemistries may generate different raw quality metrics, but clinical oncology workflows can be evaluated according to whether they produce accurate, reproducible, and timely results.

Clinical oncology NGS is not a contest over the most attractive raw sequencing metric. It is a test of whether a laboratory can produce reliable molecular answers from real-world patient specimens that guide correct treatment decisions.

For certain NGS technologies and detection chemistries, this distinction is important. Semiconductor sequencing workflows, for example, use a different sequencing principle from other sequencing chemistries, so direct comparison by Q30 alone can be technically inappropriate.

The more meaningful comparison is based on clinical performance: analytical sensitivity, analytical specificity, positive percent agreement (PPA), negative percent agreement (NPA), positive predictive value (PPV), limit of detection (LoD), reproducibility, variant-class coverage, sample compatibility, workflow success rate, and turnaround time.

When those metrics are used, the conversation becomes more relevant to clinical laboratories and procurement teams. It also becomes more aligned with what oncology testing is meant to deliver: confidence that a reported variant is real, confidence that an absent variant was not missed within the validated limits of the assay, and confidence that the result can be produced consistently in routine practice.

The practical takeaway is straightforward, even if it cuts against how many procurement documents are currently written and the marketing focus of some NGS instrument vendors. Q30 has its place as a reasonable sanity check on base-level run quality, and nobody is suggesting laboratories ignore it.

But it was never designed to answer the question that matters most in clinical oncology: did the workflow accurately find the variant, or didn't it? That question gets answered by sensitivity, specificity, LoD, reproducibility, and validated performance across the variant types the assay actually reports—not by Q-score. 

Is Q30 a bad metric?

No, but it is limited in how well it represents clinical value. Q30 can be useful for assessing base-level sequencing quality in contexts where that metric is technically appropriate. The issue is not Q30 itself. The issue is using Q30 as a universal stand-alone requirement for clinical oncology NGS workflows.

Why is Q30 not enough for targeted oncology sequencing?

Targeted oncology sequencing depends on the performance of the full workflow. Variant detection is influenced by sample quality, tumor fraction, nucleic acid input, depth of coverage, coverage uniformity, variant-calling algorithms, bioinformatic thresholds, limit of detection, and validation across intended variant classes. Q30 does not capture all of these factors.

What metric best reflects whether an assay detects true variants?

No single metric is enough, but analytical sensitivity, PPA, LoD, and reproducibility are especially important. These metrics should be evaluated for the specific variant classes the workflow is intended to detect, including SNVs, indels, CNVs, fusions, and other biomarkers where applicable.

What metric best reflects whether an assay avoids false positives?

Analytical specificity, NPA, PPV, and false-positive analysis are central. In oncology, avoiding false positives is important because a reported variant can influence interpretation, follow-up testing, trial screening, or treatment discussions.

Why does high depth matter in targeted oncology NGS?

High depth means that many reads cover the same targeted genomic region. In targeted oncology sequencing, high depth can support confident detection of low-frequency variants when combined with validated chemistry, adequate coverage uniformity, controls, and bioinformatics. Depth should not be interpreted alone, but it is an important part of the clinical-performance picture.

Are raw-read metrics more relevant for whole-genome sequencing than targeted oncology panels?

Often, yes. Whole-genome and broad research applications may place more emphasis on global base-level data quality because the analysis spans very large genomic regions. Targeted oncology panels concentrate sequencing on selected regions and should be assessed primarily by validated variant-detection performance in the intended clinical or translational context.

What should procurement teams prioritize when evaluating oncology NGS?

Procurement teams should prioritize evidence that the workflow produces accurate, reproducible, and timely variant calls for the intended specimen types and variant classes. Q30 can be reviewed where appropriate, but it should not replace sensitivity, specificity, PPA, NPA, PPV, LoD, precision, reproducibility, coverage uniformity, workflow success rate, turnaround time, and real-world implementation evidence.