Star allele calling, CPIC-guided metabolizer classification, and PGx reporting, built for labs running NGS-based pharmacogenomics programs on panels, exomes, and genomes.
Codeine: avoid use due to risk of inadequate analgesia. Consider an alternative analgesic not metabolized by CYP2D6.
VSPGx is part of VarSeq Suite, licensed for Research Use Only. It supports laboratory-developed test (LDT) development, method validation, and internal research workflows. Your laboratory is responsible for the clinical validation and regulatory compliance of any test built with VSPGx. Review the VarSeq Suite and VarSeq Dx regulatory status.
Pharmacogenomics studies how an individual's genetic variants affect their response to drugs, and it has moved steadily from research into routine testing over the past decade. The driver is straightforward: adverse drug reactions are a leading cause of preventable hospitalizations, and a meaningful share of them trace back to predictable genetic variation in drug-metabolizing enzymes and drug transporters.
The most established gene-drug interactions involve a small set of high-priority genes. CYP2D6 metabolizes roughly 25% of commonly prescribed medications, including codeine, tamoxifen, tricyclic antidepressants, and several antipsychotics. CYP2C19 governs activation of clopidogrel, the antiplatelet drug used after cardiac stent placement, so a poor metabolizer on standard dosing can have inadequate antiplatelet protection. SLCO1B1 variants predict statin-induced myopathy risk, and VKORC1 with CYP2C9 together explain most warfarin dose variability.
CPIC (the Clinical Pharmacogenetics Implementation Consortium) publishes peer-reviewed prescribing guidance for these and dozens of other gene-drug pairs, tiered by strength of evidence. FDA drug labeling now incorporates pharmacogenomic information for over 300 medications. The scientific case is established. The barrier for most labs is building the bioinformatic and interpretation infrastructure to run PGx reliably at volume. For the clinical and scientific foundations, see the pharmacogenomics guide.
PGx is technically distinct from germline or somatic variant analysis. The output is not a five-tier classification. It is a metabolizer phenotype and a drug-specific prescribing recommendation. Getting there takes several steps that standard variant interpretation tools handle poorly.
PGx genes use star allele nomenclature, where each named allele such as CYP2D6 *1, CYP2D6 *4, or CYP2C19 *17 represents a defined combination of variants with a known functional consequence. Translating raw SNV, indel, structural, and copy number calls into accurate star allele assignments is the first technical hurdle.
CYP2D6 is the hardest case: over 150 defined alleles, frequent copy number variation, and hybrid alleles formed by recombination with the adjacent CYP2D7 pseudogene. VSPGx, the dedicated PGx module in VarSeq, automates star allele identification across the clinically relevant gene set. On short-read data, VSPGx alone cannot resolve CYP2D6 because of that high homology with CYP2D7, so CypCall, the Golden Helix tool built specifically for CYP2D6 star allele assignment, handles that locus. Copy number and structural alleles called by an external tool can be incorporated through the sample manifest.
Each patient carries two copies of each gene, and the combination of two star alleles, the diplotype, determines the functional outcome. CYP2D6 *4/*4 (two non-functional alleles) produces a poor metabolizer. CYP2D6 *1/*2xN (one normal allele plus a duplicated functional allele) produces an ultra-rapid metabolizer.
VSPGx assigns diplotypes from the star allele calls and handles the combinatorial complexity of multi-allele loci, flagging cases where more than one diplotype is consistent with the observed data for review.
Diplotypes translate into metabolizer phenotypes: Poor (PM), Intermediate (IM), Normal (NM), Rapid (RM), or Ultra-Rapid (UM). VSPGx applies activity score frameworks where CPIC defines them, as for CYP2D6 and CYP2C19, and direct genotype-phenotype lookup tables elsewhere.
The result is a per-gene phenotype classification, the direct input to gene-drug interpretation, generated the same way on every run.
VSPGx links the phenotype to CPIC and FDA pharmacogenomic guidance to surface drug-specific recommendations. A CYP2D6 poor metabolizer on codeine faces inadequate analgesia, and CPIC recommends avoiding codeine in this population. A CYP2C19 poor metabolizer on clopidogrel has inadequate antiplatelet activation, and CPIC recommends alternative antiplatelet therapy.
Each recommendation surfaces at the point of interpretation, linked to the specific guideline version and evidence tier, for review by qualified laboratory personnel.
The final output is a laboratory-branded report that presents metabolizer phenotype by gene, lists current medications with gene-drug interaction alerts by severity, gives drug-specific recommendations linked to the CPIC guideline version, and summarizes the alleles tested. VarSeq supports configurable PGx report templates so labs can match institutional standards and the specific panel in scope, with interpretive comments added before sign-out by qualified laboratory personnel.
VSPGx supports the full set of CPIC Tier A and Tier B gene-drug pairs and is configurable for institutional panel designs. The genes below carry the most established prescribing implications.
| Gene | Drug Classes Affected | Key Decision |
|---|---|---|
| CYP2D6 | Opioids, antidepressants, antipsychotics, tamoxifen | Dose adjustment or drug switch for PM/UM |
| CYP2C19 | Clopidogrel, PPIs, SSRIs, antiepileptics | Alternative antiplatelet for PM; PPI dose for UM |
| CYP2C9 | NSAIDs, warfarin, phenytoin | Dose reduction for PM; warfarin bleeding risk |
| VKORC1 | Warfarin | Warfarin sensitivity; lower starting dose |
| SLCO1B1 | Statins (simvastatin, atorvastatin) | Myopathy risk; dose reduction or alternative statin |
| DPYD | Fluoropyrimidines (5-FU, capecitabine) | Severe toxicity risk in PM; contraindicated |
| TPMT / NUDT15 | Thiopurines (azathioprine, mercaptopurine) | Myelosuppression risk; dose reduction for PM |
| UGT1A1 | Irinotecan | Toxicity risk in reduced-function genotypes |
| CYP3A5 | Tacrolimus and other immunosuppressants | Dose adjustment for transplant medications |
| G6PD | Antimalarials, rasburicase | Hemolytic anemia risk; contraindicated in deficient |
Star allele definitions follow PharmVar nomenclature, and gene-drug guidance follows CPIC and FDA pharmacogenomic labeling. PharmGKB curates the underlying evidence base for these pairs.
VSPGx calls star alleles from any data source that provides genotype calls, so no PGx-only sequencing is required. For a comparison of when to choose each assay, see the panel analysis solution and the exome analysis solution.
Purpose-built panels covering 12 to 30 or more PGx genes. High-throughput, streamlined workflows for clinical volume, and the most common entry point for labs launching a PGx program.
PGx results extracted from whole exome data run for a primary indication such as rare disease or hereditary cancer. Patients gain PGx insight from sequencing already performed, and VSPGx runs in the same VarSeq session as the primary analysis.
Whole genome sequencing gives the broadest PGx coverage, including non-coding regulatory variants and structural variants that targeted panels and exomes can miss. Relevant where WGS is the primary assay and PGx is one component of the report.
For SNP microarrays that include pharmacogenomics probe sets, VarSeq calls star alleles from array data where allele coverage is sufficient for the genes in the panel.
Launching a PGx program is straightforward. Scaling it is where operational complexity emerges: maintaining interpretation consistency, standardizing reports across interpreters, and building a longitudinal patient PGx record that holds up over years.
VSWarehouse stores PGx results and diplotype classifications longitudinally, so labs can retrieve a patient's prior PGx profile when a new medication is considered. For population-scale programs, VSWarehouse supports cohort-level queries that identify every patient with a specific high-risk diplotype across the tested population. VarSeq's workflow automation runs the pipeline from sample import through star allele calling, phenotype classification, and report generation, so interpretation effort concentrates on edge cases and complex allele assignments rather than routine diplotype confirmation.
VSWarehouse stores diplotype and phenotype assignments by patient and supports cohort queries for pre-emptive PGx programs at health-system scale.
VSPipeline runs sample import through star allele calling, phenotype classification, and report generation, reducing per-sample hands-on time for high-volume labs.
Run PGx alongside rare disease, hereditary cancer, and oncology workflows on the same annotation framework, reporting infrastructure, and deployment.
PGx testing rarely runs in isolation. Labs offering hereditary cancer panels are adding PGx as a second panel, since a patient with a BRCA2 pathogenic variant being considered for olaparib also benefits from CYP2D6 metabolizer status for concurrent medications. Labs running rare disease exomes extract PGx results from the same exome data. Oncology programs add DPYD testing before fluoropyrimidine chemotherapy.
VarSeq is built for this multi-program reality. PGx analysis runs within the same platform as rare disease analysis software, precision oncology analysis, and hereditary cancer risk analysis, on one validation effort and one deployment.
Labs deploying VSPGx into a CLIA-certified workflow must validate the assay, document interpretation consistency, and standardize reports. VarSeq's deterministic pipeline produces the same star allele calls and phenotype classifications on every run, which supports validation documentation. For full lab infrastructure context, see the clinical lab infrastructure guide.
Full deployment within the institution's own infrastructure. No data leaves the institutional boundary, which suits sites with strict policies against cloud-based storage of patient data.
Deployment within the lab's own AWS or Azure environment, with administrative control, geographic data residency selection, and elastic compute scaling for variable sample volumes.
Fully offline deployment for the most sensitive environments. Software, annotation databases, and licensing operate on an isolated internal network.
Regulatory status. VSPGx is part of VarSeq Suite, licensed for Research Use Only, and supports laboratory-developed test development and method validation. It is not offered as a cleared in vitro diagnostic. Golden Helix operates under an ISO 13485-certified Quality Management System, and VarSeq releases are validated, version-controlled, and supported with change documentation for laboratory QMS integration. Review VarSeq Suite and VarSeq Dx regulatory status.
Pharmacogenomics software: what it does, how star allele calling works, and how it fits a lab's operations.
Standard germline and somatic platforms are built around variant classification, assigning one of five ACMG tiers or four AMP tiers to each variant. Pharmacogenomics needs a different output: star allele assignments, diplotype combinations, metabolizer phenotypes, and drug-specific recommendations linked to CPIC and FDA guidance. The technical work is also distinct, because genes like CYP2D6 require copy number analysis, structural variant detection, and allele phasing to call star alleles accurately, which standard variant callers are not designed to handle.
VSPGx is built for this workflow and runs inside VarSeq, so labs running both PGx and standard germline or somatic analysis do not need separate platforms.
CYP2D6 is the most technically challenging PGx gene. It has over 150 defined alleles, frequent copy number variation (whole-gene duplications producing ultra-rapid metabolizers, deletions producing poor metabolizers), and hybrid alleles formed by recombination with the adjacent CYP2D7 pseudogene. VSPGx calls SNVs and indels and assigns star alleles from the PharmVar definitions, and copy number or structural alleles called by an external tool can be incorporated through the sample manifest. On short-read data, CYP2D6 cannot be resolved reliably from VSPGx alone because of its high homology with the CYP2D7 pseudogene, so CypCall, the Golden Helix tool dedicated to CYP2D6 star allele assignment, handles that locus.
Ambiguous cases, where more than one diplotype assignment is consistent with the observed data, are flagged for review with the alternative interpretations displayed.
VSPGx covers current CPIC Tier A and Tier B gene-drug pairs. Tier A pairs have sufficient evidence for CPIC to provide specific prescribing recommendations, including CYP2D6/codeine, CYP2C19/clopidogrel, DPYD/fluoropyrimidines, TPMT and NUDT15/thiopurines, SLCO1B1/simvastatin, and others. Tier B pairs have sufficient evidence to recommend testing, with more context-dependent prescribing guidance.
FDA pharmacogenomic labeling is also integrated, covering the broader set of medications where FDA includes PGx information in the label. The gene-drug pair database is updated as CPIC publishes new or revised guidelines.
Yes, and this is increasingly common. A patient undergoing whole exome sequencing for a rare disease indication can receive a PGx report from the same data with no additional assay. VSPGx runs in the same VarSeq session as the primary analysis, applying star allele calling and phenotype classification to the PGx gene set from the exome data.
Coverage for some PGx alleles may be incomplete depending on the exome capture kit, so labs should validate per-gene coverage for their specific assay before reporting. Where coverage completeness is critical, a dedicated PGx panel or a genome assay provides more reliable allele-level coverage.
VarSeq supports configurable PGx report templates. Labs set the report to include the metabolizer phenotype for each gene tested, a medication interaction summary listing current drugs with gene-drug interaction severity, drug-specific recommendations linked to the CPIC guideline version, tested allele summaries, and laboratory interpretive comments. Report language is configurable to match institutional standards.
The goal is a report that qualified laboratory personnel can finalize and that the ordering provider can act on, with the analytical work documented for the lab's quality system.
VSWarehouse stores PGx results, diplotype classifications, and phenotype assignments longitudinally by patient. When a new medication is considered for a patient whose PGx profile is already on file, the lab retrieves the prior result rather than rerunning the assay. For health systems with population-scale pre-emptive programs, VSWarehouse supports cohort-level queries to identify patients carrying high-risk diplotypes for specific drug classes across the tested population.
This is the infrastructure layer that makes a pre-emptive PGx program sustainable at scale.
VSPGx supports star allele calling from targeted PGx panels, whole exome sequencing, whole genome sequencing, and SNP microarrays with PGx probe sets. Targeted panels are the most common entry point for labs launching a PGx program, with high throughput and purpose-built allele coverage. Exome and genome integration suits multi-indication programs where PGx is one component.
The right assay depends on the genes in scope, the required allele coverage depth, and the lab's existing sequencing infrastructure.
No. VSPGx is part of VarSeq Suite and is licensed for Research Use Only. It supports laboratory-developed test (LDT) development, method validation, and internal research workflows. It is not offered as a cleared in vitro diagnostic device, and it is separate from VarSeq Dx, which is the CE-marked product line. Your laboratory is responsible for the clinical validation and regulatory compliance of any test built with VSPGx. See the VarSeq Suite and VarSeq Dx regulatory status for details.
Request a personalized evaluation with your own pharmacogenomics data.
Deep dives into CPIC implementation, star allele calling, and PGx reporting workflows.
VSPGx is part of VarSeq Suite, licensed for Research Use Only. Not available as an in vitro diagnostic medical device. Intended for LDT development, method validation, and internal research workflows.
Star allele calling, CPIC-guided metabolizer classification, and configurable PGx report templates in one Research Use Only platform.