How we read a genome.
The pipeline is automated, validated, and the same every time, parse, quality-check, call, annotate, score. Here's what happens between your upload and your report.
Five stages, one window
Parse & QC
We detect the file format, normalise it, and run quality control, call rates, strand consistency, and build alignment, before anything else.
Variant calling
Genotypes and sequence reads are resolved into a clean set of variant calls against the reference genome.
Annotation
Each variant is annotated against curated clinical and population databases to attach meaning, function, frequency, and clinical significance.
Risk modelling
Clinical and polygenic models translate annotated variants into trait and risk interpretations with confidence context.
Integration
On the Max Plan, calls are synthesised across systems into the Gene Integration File and loaded into the AI Health Assistant.
Report
An encrypted PDF and a structured JSON/CSV export are delivered to your portal, inside the 8–12 hour window.
What we analyse against
Interpretation is only as good as the references behind it. GeneLab annotates your variants against established, curated resources spanning clinical significance, population frequency, drug response, and trait associations. Where evidence is strong, we say so; where it's preliminary, we label it that way.
- Clinical significance, curated variant–condition relationships for pathogenicity classification.
- Population frequency, allele frequencies across global reference populations, so rarity is read in context.
- Pharmacogenomics, validated drug–gene guidelines for metaboliser status.
- Trait & risk associations, genome-wide association evidence for polygenic traits and risk.
Confidence, not certainty
Genetics shifts probabilities; it rarely decides outcomes. Every risk statement in a GeneLab report carries context, what the variant does, how common it is, and how strong the evidence is. We avoid false precision, and we never present a genetic result as a diagnosis.
Quality control
Before interpretation, each file passes automated QC. Low-confidence calls are flagged rather than silently reported, and files that don't meet a minimum quality bar are returned to you with an explanation instead of a misleading report.
GeneLab is an analysis service, not a clinical laboratory diagnosis. Results are for information and education. Confirm anything clinically meaningful with a qualified healthcare professional and, where appropriate, a validated clinical test.
Read a finished report.
A full sample of what the pipeline produces.