How to Convert VCF Files into Clinical PGx Reports

To convert a VCF into a clinical PGx report, you start with already-called genotypes, load them into interpretation-support software, reconcile them against curated pharmacogenomic evidence, and let your lab's medical director review and sign out a lab-branded report. SignalPGx handles the interpretation-support and reporting steps. It intakes your called results downstream of variant and star-allele calling.

That last point is the crux of a clean VCF-to-pharmacogenomics-report pipeline, so let's make it explicit before the steps.

Where does SignalPGx start in the PGx pipeline?

SignalPGx sits downstream of variant and star-allele calling. It does not align reads, call variants from BAM or FASTQ, or perform automated star allele calling. Instead, it intakes results that are already called.

Think of the workflow in two halves. The upstream half is genotyping: your sequencer or genotyping platform produces raw data, and a bioinformatics tool for PGx genotyping data resolves that into variants and diplotypes. Common upstream callers include PharmCAT for VCF-based diplotype calling and platform-native software for Agena MassARRAY panels.

The downstream half is interpretation and reporting: taking those called diplotypes, mapping them to phenotypes, checking them against medications and guidelines, and producing a clinician-ready document. That downstream half is where SignalPGx operates. You bring your called genotypes; the software supports interpretation and lab-branded reporting.

This division matters because it means you do not need to hire a bioinformatics team or build a next-generation sequencing pharmacogenomics analysis stack from scratch to produce reports. You need reliable upstream calling, which most CLIA labs already run, plus a dependable downstream layer.

What do you need before you begin?

Before you convert anything, confirm you have three things in hand.

• Called genotype data. A VCF with resolved diplotypes, a PharmCAT output file, an Agena MassARRAY export, a structured CSV, or manually entered star alleles. If your VCF contains only raw variant positions with no diplotype resolution, run it through your star-allele caller first.
• A defined gene panel. Know which pharmacogenes your assay covers so the report scope matches your validation. SignalPGx supports interpretation across 50+ pharmacogenes.
• A licensed medical director. Final review and sign-out are performed by your lab's own medical director. The software is decision-support; it does not replace a clinician or issue reports as a service.

With those in place, the conversion itself is straightforward.

How do you convert VCF to a clinical PGx report, step by step?

Here is the practical sequence for turning a called VCF into a released, clinician-ready PGx report.

1. Export your called results. Pull the VCF, PharmCAT JSON, Agena export, or CSV from your genotyping workflow. Make sure diplotypes are resolved, not just raw variant calls.
2. Ingest the file into SignalPGx. Upload through the intake interface or an automated feed. The platform accepts VCF and PharmCAT output, Agena MassARRAY, CSV, and manual entry. See the platform overview for supported intake paths.
3. Let the engine map diplotype to phenotype. The deterministic interpretation engine acts as an automated diplotype to phenotype converter, translating your called star alleles into functional phenotypes and drawing on curated evidence.
4. Review the medication-level findings. The engine reconciles findings against 950+ medications and flags 7,700+ drugs with documented drug–drug interactions, surfacing what is clinically relevant for the report.
5. Have the medical director review and release. Your director inspects the drafted content, adjusts as needed, and signs out. Human-in-the-loop review is a core step, not an afterthought.
6. Deliver the lab-branded report. The finished document carries your lab's branding and can be delivered to the ordering clinician directly or through your existing systems.

Each step keeps the boundary clean: the software supports interpretation and formatting; your director owns the clinical sign-out; the treating physician owns the prescribing decision.

What happens during interpretation support?

The interpretation step is where a pile of diplotypes becomes something a clinician can act on. SignalPGx runs a deterministic interpretation engine that reconciles 16 evidence sources into a single, consistent view called the Medication Intelligence Graph.

Those 16 sources are FDA, OpenFDA, HCSC (Health Canada), EMA, Swissmedic, PMDA, CPIC, DPWG, PharmGKB, PharmCAT, PharmVar, ClinVar/ClinGen, RxNorm, DailyMed, Onsides, and gnomAD. Reconciling them by hand for every report would be slow and error-prone; doing it deterministically means the same input produces the same, traceable output every time.

You can read more about how these sources are combined on the Medication Intelligence Graph page. For guideline context on how phenotypes map to prescribing recommendations, the CPIC guidelines are the standard reference many labs build against, and PharmVar is the authority for star-allele nomenclature your upstream caller should follow.

If your team wants to ask questions about a specific finding, the guardrailed SignalAI assistant answers with citations or declines when it cannot cite a source. It supports your reviewers; it never makes autonomous clinical decisions. Details are on the SignalAI page.

How does the medical director review and release the report?

Interpretation support produces a draft; it does not produce a released report on its own. Your lab's licensed medical director is the human in the loop who reviews the drafted interpretation, confirms it against the patient context, and signs out.

This structure keeps the regulatory lines clear. CLIA certification belongs to your laboratory, not to the software. SignalPGx is reporting and decision-support software; it is not a diagnostic test and is not FDA-cleared or approved. It helps your director work faster and at scale, but the clinical judgment and the sign-out remain theirs, and the final prescribing decision remains with the treating physician.

How do lab-branded reports and integrations fit in?

Once a report is signed out, delivery should match how your lab already operates. SignalPGx produces white-label reports carrying your name, logo, and formatting, so the clinician sees your lab's brand, not a third-party vendor.

For distribution, the platform supports FHIR, SMART-on-FHIR, and CDS Hooks, plus LIS integration, so reports and structured results can flow into the systems clinicians already use. The integrations page covers the available connection paths, and the PGx reporting overview walks through the end-to-end reporting product.

Billing support is built in as well, including MolDx Z-code, CPT, and ICD-10 handling, so the reporting workflow connects to reimbursement rather than sitting apart from it.

What keeps reports current after they are released?

Pharmacogenomic guidance is not static. When a guideline moves, a report you issued last quarter may no longer reflect the latest evidence. That is a real clinical-hygiene problem for any lab producing PGx reports at volume.

SignalPGx addresses it with living reanalysis. The platform re-evaluates previously issued reports when the underlying guidelines change and flags affected reports for your medical director to review. It does not silently reissue anything; a human decides what to do with each flag. See living reanalysis for how the re-evaluation and flagging work.

How do you get started converting VCFs into reports?

If your lab already produces called genotypes from VCF, PharmCAT, or Agena workflows, the path to clinician-ready reports is short: connect your intake, let the engine support interpretation, have your director sign out, and deliver a lab-branded report. There is no bioinformatics hire and no build cost required to stand up the downstream layer.

Per-report pricing across the All Signal, Scale, and Enterprise tiers is outlined on the pricing page. To walk through your specific intake formats and panel, reach the team through contacts, and browse more practical guides on The Signal.

The short version: keep your trusted upstream calling, and let SignalPGx handle interpretation support and lab-branded reporting downstream of it.