Health Hub · the platform
From raw reads to a queryable cohort.
The same Health Hub platform that runs the clinical agents runs genomics: a reproducible pipeline with the QC discipline research demands, an output layer that links sequence to clinical phenotype, all on the Forge OS.
Raw reads → linked records → cohort → queryable
End to end
From sample registration to a signed report.
One reproducible pipeline carries every sample from accessioning through interpretation, secondary and tertiary analysis, review and reporting, in one place.
- 01
Sample registration & tracking
Barcoded accessioning and full chain-of-custody, every sample tracked from receipt to result.
- 02
Secondary analysis
Alignment, joint calling and variant calling (SNVs, indels and CNVs) on a documented, reproducible pipeline.
- 03
Quality control
Sample- and variant-level QC with MultiQC, captured in a detailed, defensible QC report.
- 04
Annotation
Functional, population and clinical annotation, ready to interpret.
- 05
Tertiary analysis
Variant filtering and prioritisation, AI-driven ACMG / AMP classification, and automated variant-description generation.
- 06
Review & reporting
An embedded IGV variant viewer, clinical history mapped to HPO, and AI-generated, sign-off-ready reports.
Sample registration through secondary and tertiary analysis to a signed report, one pipeline.
Interpretation, accelerated
The tertiary layer, built in.
Everything an analyst needs to go from a VCF to a defensible call, AI-assisted, evidence-cited, human-confirmed.
AI-driven ACMG classification
Variants classified to ACMG / AMP criteria with cited evidence per rule, the analyst reviews and confirms.
Filtering & prioritisation
Narrow millions of variants to the handful that matter, by phenotype, inheritance, frequency and predicted impact.
IGV variant viewer
Inspect any call in an embedded IGV viewer, reads, coverage and context, right next to the evidence.
Clinical history → HPO
Free-text clinical notes mapped to HPO terms, driving phenotype-aware filtering and prioritisation.
Variant description generation
Auto-drafted, reviewable variant descriptions and interpretations, consistent and citation-backed.
Automated AI reporting
Structured, on-template reports assembled from the evidence, generated automatically, ready to sign.
Your data, in full
The only provider that hands back everything.
Most labs return a PDF. We return the complete data estate (every intermediate file, the full QC, and any custom output you ask for) so the science is yours to re-analyse, audit and build on.
FASTQ
BAM
CRAM
VCF
Annotated VCF
CNV calls
MultiQC reports
Detailed QC report
Custom data on request
The infrastructure
Built to live next to clinical data.
FHIR-linked
Genomic results link to clinical phenotype in FHIR R4, so sequence and outcome sit in one record.
On-prem capable
Sequencing runs at accredited partner labs; the entire analysis pipeline can run inside your institution, so genomic data never has to leave.
Any cloud
Or run it on the cloud you already use, with the same reproducible workflow.
Reproducible QC
Documented, repeatable quality control, the difference between data and a defensible cohort.
Same platform as clinical
The orchestration layer is shared with the clinical agents, one company, one engine.
Cohort-ready
Outputs are structured for gene-burden and association analyses out of the box.
Bring us your samples or your questions.
Whether it's a single cohort or standing infrastructure, we'll scope the pipeline to fit.