Analytical & Clinical Validation

A platform engineered for regulatory, payer, and clinical review.

InterceptIQ™ has been validated across every category of analytical performance and benchmarked against prospective clinical cohorts. The data below summarize the studies that anchor the platform's scientific credibility.

Performance scorecards

Headline performance across every validation pillar.

CLIA / CAP / NYSDOH · Method comparison vs. orthogonal ddPCR

Linearity (R²)PASS
0.9999

0.5 – 5,000 cp/µL

Limit of DetectionPASS
4.5cp/µL

95% probability · Probit

Diagnostic AUCPASS
0.979

T1D vs. control · n = 1,847

Inter-run CVPASS
3.2%

Across 28 days · 4 operators

SpecificityPASS
97.8%

At calibrated cutoff

SensitivityPASS
94.6%

At calibrated cutoff

Validation framework

Eight pre-defined acceptance criteria. All met or exceeded.

The InterceptIQ™ assay was validated in accordance with CLSI guidance and internal protocols developed for regulatory submission. Acceptance criteria were pre-specified before unblinding to preserve statistical integrity.

Accuracy

Bias < 2% across operating range

Precision

Inter-run CV 1.6% – 5.4%

Linearity

R² = 0.9999 · 4-log dynamic range

Specificity

No cross-reactivity in 96-sample panel

Recovery

98.1% – 102.4% spike recovery

Stability

≥ 7 days · plasma at 4 °C

Limit of Detection

4.5 copies / µL · 95% Probit

Reference Range

Established across 3 cohorts (n = 1,847)

Analytical performance

Linearity, precision, and limit of detection — measured at scale.

Analytical linearity

R² = 0.9999 · 0.5 → 5,000 copies / µL

PASS
1,2502,5003,7505,00001,2502,5003,7505,000Measured (cp/µL)Expected (cp/µL)
Slope
1.0021
Intercept
−0.34
0.9999

Precision · inter-run CV

CV < 6% across operating range

PASS
Low (10 cp/µL)CV 5.4%
Mid (100 cp/µL)CV 3.2%
High (1,000 cp/µL)CV 2.1%
V. High (5,000 cp/µL)CV 1.6%
n = 240 replicates · 4 operators · 3 reagent lots · 28 days Acceptance: 6%

Limit of detection · Probit

LoD95 = 4.5 copies / µL

PASS
25%50%75%95%4.5 cp/µL05101520Input concentration (cp/µL)

Diagnostic performance · ROC

AUC = 0.979 · T1D vs. control

PASS
Operating point · Sn 94.6 / Sp 97.81 − SpecificitySensitivity

Clinical reference ranges

Established stratification across healthy, at-risk, and active cohorts.

Reference ranges were established prospectively across three IRB-approved cohorts with longitudinal sampling. The calibrated cutoff yields 94.6% sensitivity and 97.8% specificity at the operating point.

Clinical reference ranges

Stratification across cohorts (n = 1,847)

ESTABLISHED
Clinical cutoff · 280255075100InterceptIQ™ score
Healthy control
12 ± 7
At-risk (Stage 1–2)
38 ± 10
Active T1D
72 ± 11

Multi-site methylation architecture

Three independent epigenetic signals.
One calibrated intelligence score.

Rather than relying on a single biomarker, InterceptIQ™ interrogates multiple tissue-specific methylation loci across the insulin gene. The joint signal raises specificity, lowers false-positive rate, and produces a clinically interpretable disease intelligence output.

INS · Chromosome 11p15.5 · bisulfite-converted track
Δ from TSS (bp)
−500−250TSS+250+500
Exon 1Exon 2Exon 3TSSINS -233upstreamINS -135proximalINS +399intragenic
INS -233w = 0.34
81%
Unmethylated · β-cell

Open chromatin in pancreatic β-cells; hypermethylated in non-β tissue.

Promoter · upstream
INS -135w = 0.33
74%
Unmethylated · β-cell

β-cell specific demethylation; conserved across human islet donors.

Promoter · proximal
INS +399w = 0.33
69%
Unmethylated · β-cell

Independent confirmatory locus; lowers false-positive rate vs. single-site assays.

Exon 2 · intragenic
InterceptIQ™ score
74.7
Joint disease intelligence

Calibrated weighted combination — auditable, interpretable, and reproducible across cohorts.

Calibrated · AUC 0.979 · n = 1,847

Methods & transparency

Reproducibility is part of the product.

Pre-registered protocols

Acceptance criteria, statistical analysis plans, and stopping rules are pre-registered before sample analysis.

Containerized pipelines

Every analytical run executes inside a versioned, hash-pinned container, enabling third-party re-execution.

Orthogonal confirmation

Methylation calls are cross-validated against ddPCR and an independent whole-genome bisulfite reference.

Open methods

Sequencing chemistry, basecaller version, and reference atlas builds are documented in the Publication Center.