Purpose-built for clinical laboratories diagnosing rare and undiagnosed genetic disease. Trio analysis, HPO-driven variant prioritization, ACMG-guided classification, and lifelong VUS reclassification monitoring in one validated platform.
Rare disease patients wait an average of 5 to 7 years for a molecular diagnosis. They see eight or more physicians. They receive two or three misdiagnoses along the way. And at the end of that journey, the most common outcome from genomic sequencing, for roughly half of all patients tested, is not a definitive answer. It is a variant of uncertain significance.
The diagnostic odyssey is not primarily a sequencing problem. Sequencers are fast, accurate, and affordable. The bottleneck is interpretation: reducing tens of thousands of variant calls to a defensible clinical conclusion, efficiently enough to return results within a clinically meaningful timeframe, consistently enough to hold up under CAP/CLIA inspection, and with enough institutional memory to revisit yesterday's uncertain result when today's evidence reclassifies it.
This is the problem rare disease analysis software exists to solve. It is also the problem that separates platforms built for clinical production from tools built for research exploration. For a deeper look at the rare disease diagnostic journey, see the rare disease genomics guide.
Mendelian traits cataloged to date
Integrated SNV & CNV detection workflow
Trio WES diagnostic yield in unselected rare disease cohorts
Rare diseases, while individually rare, collectively affect more than 300 million people worldwide. Roughly 80% have a genetic cause. Next-generation sequencing now assesses thousands of conditions in a single assay, but yield depends on what the lab does with the resulting variant call file.
Moving from gene panels to whole exome or whole genome sequencing raises the probability of finding the causal variant, especially when paired with native trio analysis.
Replace sequential single-gene tests, microarray, and MLPA with one NGS workflow that calls SNVs, indels, and CNVs in a single clinical pipeline.
A molecular diagnosis guides medical management, informs recurrence risk for families, and connects patients to specialty referrals and clinical trials.
Generic NGS analysis software handles variant calling and ACMG classification. Rare disease software does more. It is built for the specific analytical, clinical, and operational demands of diagnosing Mendelian disease at clinical scale.
The single most impactful analytical capability for rare disease diagnosis is trio sequencing: analyzing the proband alongside both biological parents at once. Trio analysis detects de novo variants directly, confirms compound heterozygous phase, and produces segregation evidence that pushes a VUS toward Likely Pathogenic.
Published studies show 10–15 percentage point higher diagnostic yield for trio WES versus proband-only WES. VarSeq treats trios as a first-class workflow with automated de novo detection, inheritance model filtering, and phased compound het confirmation across panels, exomes, and genomes.
A whole exome run produces 25,000 to 40,000 variant calls. After population frequency filtering and quality thresholds, a clinical geneticist may still face hundreds of candidates. Without phenotype-driven prioritization, finding the causal variant is a manual exercise that does not scale.
VarSeq's PhoRank algorithm scores each candidate gene by how well its known disease phenotype matches the patient's HPO terms. A variant in a gene with Definitive ClinGen classification for the patient's exact feature set surfaces to the top of the review queue automatically. A variant in a gene with no phenotypic relevance moves down the list without being discarded.
Manually applying the 28 ACMG/AMP criteria across hundreds of candidate variants per case is the primary bottleneck in rare disease interpretation. It introduces reviewer-to-reviewer inconsistency, extends turnaround time, and creates documentation gaps that surface during CAP inspection.
Automated criteria scoring calibrated to current ClinGen gene-specific specifications, applied to every variant in every case, cuts per-case interpretation time substantially while improving classification consistency. Reviewers validate the automated assessment, adjust criteria where professional judgment is required, and sign out a fully documented classification with complete audit trail.
VUS rates of 30 to 50% are typical in unselected WES populations. A VUS at the time of reporting is not the end of the diagnostic process. It is the start of an evidence-accumulation process that may take months or years to resolve. Published studies document VUS reclassification rates of 10 to 30% over 3 to 5 years.
Rare disease software must maintain a living connection between each patient's variant history and the evolving state of external classification databases, alerting clinical staff when ClinVar or other authoritative sources reclassify a variant previously reported to a patient. This is a patient safety function, not an optional feature.
Reanalysis of previously negative WES data yields new diagnoses in 10 to 20% of cases when performed 2 to 3 years after the initial test, not because the sequencing was inadequate, but because new gene-disease relationships were established in the interim. The variant was always there. The knowledge to interpret it was not. A rare disease platform must support systematic reanalysis: identifying patients whose prior results should be reconsidered in light of new evidence, and enabling efficient reinterpretation without repeating the full analysis workflow from scratch.
Trio and extended family analyses are the gold standard for rare disease investigation. VarSeq compares the proband's variants against parental genotypes and surfaces the most likely pathogenic candidates with inheritance-informed filter chains.
Flexible pedigree support from single proband through parent-child duos, full trios, and quads with automated de novo, compound heterozygous, and homozygous recessive filtering.
Model complex inheritance patterns for extended families, including consanguineous pedigrees and multi-generational studies.
Visually confirm variant segregation across affected and unaffected family members to build evidence for pathogenicity.
Automated identification of variants matching expected Mendelian inheritance models (AD, AR, XL, mitochondrial).
A complete diagnosis requires the full spectrum of variation. VarSeq integrates SNVs, indels, and CNVs into a single guided ACMG workflow.
Automate the scoring of germline variants using the ACMG five-tier framework. VSClinical aggregates population data, functional predictions, and literature evidence for rapid classification.
Clinical InterpretationFirst to commercialize the ACMG/ClinGen CNV guidelines. Score gains and losses alongside small variants in a unified interface, including family-based segregation analysis.
CNV AnalysisVisualize variants in their genomic context with integrated Golden Helix GenomeBrowse. Cross-reference master literature databases like Genomenon to surface clinical evidence.
Golden Helix GenomeBrowseVSWarehouse is the enterprise data layer that converts individual case results into cumulative institutional knowledge. It is also the reclassification monitoring engine that keeps prior results current as external evidence evolves.
Every variant classified in VarSeq is stored in a shared institutional knowledgebase. When the same variant appears in a future case, the prior assessment, evidence, and reviewer notes surface immediately. No duplicate interpretation work.
Variant frequencies across the laboratory's full patient cohort are tracked automatically. Variants enriched in the internal population may represent platform-specific artifacts or population-specific variants underrepresented in gnomAD. Both improve clinical accuracy over time.
VSWarehouse monitors ClinVar and other authoritative sources continuously and alerts clinical staff when a variant previously reported to a patient is reclassified. Alerts include prior classification, new classification, source, and a direct link to the relevant case for re-review.
Bidirectional integration through documented API connectors and HL7 FHIR messaging for automated sample tracking, result delivery, and structured variant export.
Published clinical cohort studies give the most reliable benchmarks for rare disease diagnostic yield by assay and patient population. The interpretive infrastructure a lab deploys, automated phenotype prioritization, current annotations, and systematic reanalysis, each contributes independently to the fraction of patients who reach a molecular diagnosis.
| Test Type | Typical Diagnostic Yield |
|---|---|
| Proband-only WES, unselected rare disease | 20–30% |
| Trio WES, unselected rare disease | 35–50% |
| WGS, WES-negative patients | 25–35% |
| WGS, first-line undiagnosed disease | 35–55% |
| WES reanalysis, previously negative cases | 10–20% additional |
| Rapid WGS, critically ill neonates | 30–50% |
Yield figures: Lionel et al. (2018) Genetics in Medicine; Stark et al. (2017) JAMA Pediatrics; Retterer et al. (2016) Genetics in Medicine; Wright et al. (2023) NEJM, Deciphering Developmental Disorders (DDD) study. Individual lab yield depends on patient selection, phenotype documentation quality, and the analytical approach, including whether trio analysis is performed and whether systematic reanalysis is conducted on previously negative cases.
Whether you are running targeted gene panels, whole exomes, or whole genomes, VarSeq provides a scalable infrastructure for high-throughput rare disease diagnosis. For a deeper comparison of when to choose each assay, see the WES vs WGS guide.
High-depth sequencing of defined gene sets for specific clinical indications, with configurable filter chains, ACMG classification, and clinical reporting for panels from a few genes to several hundred.
The diagnostic workhorse for rare disease. Proband-only, duo, and trio workflows with integrated CNV detection, secondary findings, and phenotype-driven prioritization.
The most comprehensive view. Detect non-coding variants, structural events, and repeat expansions that exome capture misses, with phenotype-driven prioritization across the full genome.
Rare disease patient data is PHI subject to HIPAA, GDPR for EU patient populations, and CAP/CLIA accreditation requirements. Golden Helix supports three deployment models so labs can match the data sovereignty and security requirements of any clinical environment. 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. Complete data sovereignty for sites with strict security policies or policies against cloud-based PHI storage.
Deployment within the lab's own AWS or Azure environment. Full administrative control, geographic data residency selection for GDPR compliance, and elastic compute scaling for variable sample volumes.
Fully offline deployment for the most sensitive environments. All software, annotation databases, and licensing operate on an isolated internal network.
Quality system. Golden Helix operates under an ISO 13485-certified Quality Management System. All VarSeq releases are validated, version-controlled, and supported with change documentation designed for laboratory QMS integration and CAP/CLIA inspection readiness. VarSeq Dx is CE marked under IVDR 2017/746.
Rare disease analysis software: what it is, how it improves yield, and how it fits clinical lab operations.
Rare disease analysis software is a clinical bioinformatics platform built for diagnosing rare and undiagnosed genetic conditions through genomic sequencing. It covers the full tertiary analysis workflow: variant annotation, filtering, phenotype-driven prioritization, ACMG/AMP classification, and clinical report generation, with capabilities optimized for rare disease: native trio analysis for de novo detection, HPO-based phenotype matching, VUS management, and proactive reclassification monitoring.
Unlike general-purpose NGS tools, rare disease software is built for Mendelian disease interpretation, where inheritance pattern, phenotypic specificity, and the evolving state of gene-disease knowledge are central to every classification decision.
Through several complementary mechanisms. Automated phenotype prioritization using HPO terms surfaces variants in phenotypically matched genes before manual review begins, reducing the risk of missing a relevant finding in a large candidate list. Native trio analysis enables de novo detection and compound heterozygosity confirmation that substantially improves yield over proband-only analysis. Current annotation databases, updated monthly rather than annually, ensure classification draws on the most recent ClinVar evidence. Systematic reanalysis lets labs revisit previously negative cases as new gene-disease relationships are established, recovering diagnoses from data already generated.
Together these capabilities account for the documented 10 to 20% additional yield that systematic reanalysis produces over initial analysis alone.
A targeted gene panel sequences a predefined set of genes relevant to a specific clinical indication, for example a neuromuscular or epilepsy panel. It delivers high depth per gene, fast turnaround, and a manageable interpretation scope, but it misses any causal variant outside the panel's gene set.
Whole exome sequencing covers all ~20,000 protein-coding genes in a single assay, capturing an estimated 85% of known disease-causing variants. WES is the better choice when the clinical presentation is broad, the differential spans many possible genetic causes, or prior panel testing has been non-diagnostic. WGS extends coverage to non-coding regions, structural variants, and repeat expansions and detects the additional 15 to 25% of diagnoses that WES misses in undiagnosed patients. For a detailed comparison, see the WES vs WGS guide.
A VUS means current evidence is insufficient to classify the variant as definitively pathogenic or benign. It should not drive clinical management changes. The appropriate response runs three parallel tracks: document the finding and counsel the patient about its uncertain status; pursue additional evidence where feasible, including functional studies, segregation in additional family members, or referral to specialist centers; and establish proactive reclassification monitoring so the lab is alerted when external databases update the variant's classification.
VUS reclassification rates of 10 to 30% over 3 to 5 years are well documented, with the majority resolving to Likely Benign or Benign. A smaller but clinically important fraction reclassify upward to Likely Pathogenic. That is why reclassification monitoring is a patient safety function, not an optional administrative step.
Yes. VarSeq supports WGS tertiary analysis with non-coding variant annotation, structural variant interpretation, repeat expansion analysis, and phenotype-driven prioritization across the full genome.
WGS interpretation needs analytical capabilities beyond WES: non-coding annotation sources (ENCODE regulatory elements, SpliceAI deep intronic predictions, GTEx expression data), structural variant caller integration, and tiered analysis frameworks that begin with coding regions before expanding to non-coding candidates. VarSeq handles all of these inside one validated environment, without requiring separate tools for different variant classes or genomic regions.
Trio sequencing, analyzing the affected patient alongside both biological parents, is natively supported in VarSeq as a first-class workflow, not a post-processing step. When parental samples are loaded alongside the proband, VarSeq applies inheritance model filters that identify de novo variants (present in the child, absent from both parents), confirm compound heterozygous phase (one variant from each parent), and filter homozygous variants consistent with autosomal recessive inheritance.
These inheritance-informed filter chains substantially reduce the candidate variant list relative to proband-only analysis and provide direct evidence that contributes to ACMG criteria scoring, particularly PS2 (confirmed de novo) and PM3 (detected in trans with a known pathogenic variant).
Yes. VSWarehouse provides bidirectional integration with laboratory information management systems and electronic health records through documented API connectors and HL7 FHIR messaging. Integration enables automated sample tracking from LIMS order entry through VarSeq analysis completion, automated result delivery to the EHR, and structured variant data export for downstream clinical decision support.
VarSeq also supports role-based access control integration with institutional identity management including SAML, LDAP, and Active Directory for automated user provisioning and deprovisioning. Implementation support for LIMS and EHR integration is available through Golden Helix's professional services team.
Explore featured articles and expert-led webcasts on clinical rare disease diagnosis.
Trio analysis, HPO-driven prioritization, ACMG classification, and proactive VUS reclassification monitoring in one CAP/CLIA-ready platform.