Golden Helix · Clinical Genomics Guide

Whole Genome Sequencing (WGS)
The Most Comprehensive Genetic Test

A clinical guide to WGS for clinicians, families, and lab professionals. Coding and non-coding variants, structural variants, repeat expansions, rapid WGS in critical care, and what running it actually takes.

3.2B Base PairsNon-Coding VariantsStructural VariantsRapid WGSMitochondrial

Introduction

Nothing is excluded by design.
Every gene, every regulatory region, every variant class.

Whole genome sequencing is the most comprehensive genetic test available in clinical medicine today. It sequences all 3.2 billion base pairs of the human genome in a single assay. For decades, genetic testing meant choosing between limited options: a single gene, a targeted panel, or the protein-coding exome. Each choice was a hypothesis about where the causal variant might be hiding. WGS removes that bet.

The technology is no longer experimental. WGS is being ordered for critically ill newborns in NICUs where a molecular diagnosis in 24 hours can change whether a child lives or dies. It is being used as a first-line test for undiagnosed rare disease in national genomics programs across the UK, Australia, and the US. And it is increasingly supported by professional society guidelines as the preferred approach when prior testing has failed to provide answers.

~3.2B
Base pairs sequenced per WGS sample
~4–5M
Variant calls produced per WGS sample
30–40x
Standard clinical WGS mean coverage
29%
WGS diagnostic yield in WES-negative families (NEJM 2023)
24–72h
Rapid WGS turnaround for critically ill neonates
$200–$400
Research-scale WGS per sample on NovaSeq X

Definition

What Is Whole Genome Sequencing?

WGS determines the complete DNA sequence of an individual's genome: ~3.2 billion base pairs across 23 pairs of chromosomes, plus the mitochondrial genome. Unlike WES, which captures only the 1 to 2% that codes for proteins, WGS sequences everything: coding regions, introns, regulatory sequences, repetitive elements, and the vast non-coding landscape that makes up 98% of the human genome.

This comprehensiveness is both WGS's greatest strength and its greatest analytical challenge. A single WGS run produces approximately 4 to 5 million variant calls per sample, compared to 25,000–40,000 for WES. The clinical value lies not in generating those variants but in having a validated pipeline and interpretation platform capable of reducing them to a clinically actionable shortlist.

WGS is performed without an exome capture step. The entire genome is sequenced directly, without first enriching for specific regions. This eliminates the systematic coverage gaps and GC-content bias introduced by hybridization capture, and it is what allows WGS to detect variant classes that other sequencing approaches fundamentally cannot.

The Pipeline

From Sample to Report

WGS shares the same overall structure as the WES pipeline, but with critical differences at every stage: no capture step in library prep, much higher data volumes, and a more complex tertiary analysis to manage.

  1. 01

    Sample collection & DNA extraction

    Peripheral blood is standard. Saliva or buccal swabs are used when blood is not accessible. For cancer applications, tumor tissue (fresh frozen or FFPE) is sequenced alongside a matched germline sample. Input DNA quality requirements are less stringent than WES because there is no capture step depending on probe hybridization.

  2. 02

    Library preparation

    DNA is fragmented to ~150–400 bp and adapters are ligated. No enrichment step. No capture hybridization, no probe set to optimize, no kit-specific coverage variability to manage. Simpler than WES library prep.

  3. 03

    Sequencing (short-read or long-read)

    Standard clinical WGS uses short-read platforms (primarily Illumina) at 30–40x coverage with 150 bp paired-end reads. Long-read platforms (Oxford Nanopore, PacBio) read individual DNA molecules without fragmentation, producing reads of thousands to hundreds of thousands of base pairs. Output: a FASTQ file of ~100–200 GB.

  4. 04

    Secondary analysis: alignment & multi-caller variant calling

    Reads map to GRCh38 with BWA-MEM2 or Sentieon BWA, producing an 80–120 GB BAM file. Multiple callers run in parallel: GATK HaplotypeCaller or Sentieon DNAscope for SNVs/indels, CNV callers across read-depth, Manta or DELLY for structural variants, ExpansionHunter for repeat expansions, a separate pipeline for mitochondrial variants. Integration into a unified variant set is the defining secondary-analysis challenge of WGS.

  5. 05

    Tertiary analysis: tiered interpretation

    Most clinical WGS uses a tiered approach: first the coding ACMG workflow (since 85% of known disease-causing variants are still in the exome), then non-coding variants when coding is unrevealing. Virtual gene panels limit initial analysis to genes associated with the phenotype. Non-coding annotation needs ENCODE regulatory data, GTEx expression, SpliceAI, and conservation scores. See the tertiary analysis guide for depth.

Capabilities

What WGS Detects

WGS detects all major classes of genetic variation in a single assay. This is its defining clinical advantage over every other sequencing approach.

Coding region

SNVs in all 20K genes

Comparable sensitivity and specificity to WES for coding regions, and superior to WES in regions WES captures poorly: high-GC exons, first and last exons, regions with high homology to other genomic loci.

Small variants

Insertions & Deletions

Indels up to ~50 bp detected genome-wide. More uniform than WES indel calling because coverage is not distorted by capture efficiency variation.

Copy variation

Copy Number Variants

More sensitive and uniform than WES-based CNV calling because whole-genome coverage is not subject to capture probe density variation. Detects both large CNVs (matching chromosomal microarray) and smaller intragenic CNVs microarray misses.

Genome architecture

Structural Variants

Inversions, translocations, large deletions, large insertions, complex rearrangements. One of WGS's biggest wins: the capture step in WES disrupts the read-pair architecture SV callers depend on, making WES essentially blind to balanced structural variants. Gene fusions (BCR-ABL1, EML4-ALK) detectable without a separate targeted assay.

Trinucleotide & STR

Repeat Expansions

Detection at known loci with ExpansionHunter or STRipy. Critical for Huntington disease, fragile X, myotonic dystrophy, Friedreich's ataxia, spinocerebellar ataxias. Long-read WGS substantially improves sensitivity and resolves complex repeat structures.

Inheritance

Mitochondrial Genome

Complete mtDNA coverage at very high depth (mtDNA is present in hundreds of copies per cell, so relative depth is naturally elevated). Enables sensitive detection of mitochondrial variants and heteroplasmy. WES mtDNA coverage is partial and inconsistent.

Regulatory + intronic

Non-Coding Variants

The capability no other sequencing approach can replicate. Promoters, enhancers, deep intronic splice-affecting variants, 5' UTRs creating upstream open reading frames. 15 to 25% of WES-negative Mendelian disease patients are eventually diagnosed from non-coding variants detectable only by WGS.

Test Ordering

When to Order WGS

  • 01

    Undiagnosed rare disease or WES-negative patients

    The highest-yield application of WGS. A 2023 NEJM study demonstrated 29.3% diagnostic yield in WES-negative rare disease families. Where WGS is accessible and cost-effective, it can replace the stepwise testing paradigm (panel, then WES, then WGS) with a single comprehensive first-line assay.

  • 02

    Critically ill neonates and children, rapid WGS

    Rapid WGS returning results in 24 to 72 hours is one of the most evidence-supported applications in clinical genomics. See the dedicated section below.

  • 03

    Suspected structural variant etiology

    Unexplained intellectual disability with normal chromosomal microarray, complex genomic rearrangements, or phenotypes suggesting chromothripsis. Where the suspicion specifically involves a balanced rearrangement, inversion, or complex SV, WGS is the appropriate first-line molecular test.

  • 04

    Repeat expansion disorders

    When a repeat expansion disorder is in the differential (particularly for ataxia, myopathy, or neuropathy presentations where the specific locus is unknown), WGS with repeat expansion analysis is more efficient than sequential single-locus testing.

  • 05

    Mitochondrial disease

    WGS provides both complete mitochondrial genome sequencing and nuclear genome coverage for the 300+ nuclear-encoded genes involved in mitochondrial function, in a single assay.

  • 06

    Cancer genomics (tumor-normal paired WGS)

    Comprehensive cancer profiling in a single test: somatic SNVs, indels, CNVs, structural variants, mutational signatures, and TMB. For research and advanced clinical programs. See somatic variant analysis for the full clinical treatment.

Critical Care

Rapid WGS in the NICU and PICU

Rapid WGS, achieving results in 24 to 72 hours rather than the standard 3 to 8 weeks, is one of the most significant developments in clinical genomics of the past decade. Its primary application is in neonatal and pediatric intensive care, where time to diagnosis directly affects clinical management and outcomes.

Why it matters

Approximately 35% of NICU admissions have an underlying genetic etiology. For these infants, empirical management without a molecular diagnosis is often ineffective, potentially harmful, and always expensive. A molecular diagnosis can change clinical management: stopping ineffective treatments, initiating targeted therapies, guiding surgical decisions, or enabling palliative care discussions grounded in prognosis rather than uncertainty.

The evidence base

  • Project Baby Bear (California, 2021): Randomized controlled trial across five California children's hospitals showed rapid WGS changed clinical management in 19 of 184 enrolled infants (10% management change rate), with reduced costs of care and shorter hospital stays for diagnosed patients.
  • NIH NICU Study (JAMA Pediatrics, 2021): Randomized clinical trial demonstrated rapid WGS significantly improved diagnostic yield in critically ill neonates: 30% vs 17% diagnostic rate at 28 days.
  • Across multiple studies: rapid WGS consistently achieves diagnostic yields of 30 to 50% in critically ill infants with suspected genetic disease, with TAT of 24 to 72 hours for ultra-rapid programs.

What rapid WGS requires

Ultra-rapid WGS is not standard WGS performed faster. It requires purpose-built infrastructure at every stage: streamlined library prep completed in 5 to 8 hours, dedicated sequencing capacity without batching delays, optimized secondary analysis pipelines completing in hours, automated tertiary analysis with pre-populated interpretation templates for common neonatal genetic conditions, and 24/7 reporting capability with on-call clinical genetics staff. Standard clinical WGS infrastructure is not configured for this. Rapid WGS programs require deliberate investment in people, pipeline, and process.

Evidence

Diagnostic Yield by Indication

WGS diagnostic yield varies by indication and patient selection. These estimates are drawn from published clinical cohort studies. WGS consistently achieves 8 to 15 percentage points higher yield than WES for the same populations, driven by non-coding variant detection, superior SV calling, and uniform coverage of difficult exonic regions.

Clinical IndicationWGS Diagnostic Yield
Undiagnosed rare disease, WES-negative25–35%
Undiagnosed rare disease, first-line WGS35–50%
Critically ill neonates (rapid WGS)30–50%
Developmental delay / intellectual disability40–55%
Epilepsy35–50%
Multiple congenital anomalies45–60%
Suspected mitochondrial disease35–50%

For a detailed comparison of yield by indication for WGS vs WES, see the WES vs WGS guide.

The Real Challenge

Interpretation at Genome Scale

A single WGS run produces approximately 4 to 5 million variant calls. After quality and population frequency filtering, a clinical genome still contains hundreds of thousands of variants requiring analytical processing. Orders of magnitude more than a clinical exome.

This is why WGS interpretation is genuinely harder than WES interpretation, and why the platform a lab chooses for tertiary analysis has a larger impact on WGS diagnostic yield than on WES yield.

WGS-specific challenges

  • Non-coding annotation requires databases most WES pipelines do not include: ENCODE regulatory annotations, GTEx tissue-specific expression, SpliceAI, conservation scores calibrated for non-coding regions. Without these, non-coding analysis is either skipped or produces unmanageable noise.
  • Multi-caller integration: SNV/indel callers, SV callers, CNV callers, repeat expansion callers integrated into a unified, non-redundant variant set. Custom bioinformatics work without a platform designed for it.
  • Storage and compute at scale: 80–120 GB of BAM per sample, 500 MB–1 GB of VCF. At 50 samples/week, 4–6 TB of new BAM data weekly plus archival storage. See the clinical lab infrastructure guide for depth.
  • Turnaround time pressure: larger data means every pipeline stage takes longer. Without optimized parallel pipelines, WGS TAT slips from days to weeks, unacceptable for clinical programs.

Labs that implement WGS successfully invest in a platform purpose-built for genome-scale analysis, not an exome pipeline adapted to handle larger inputs.

Economics

Cost and Insurance Coverage

Clinical WGS pricing inclusive of interpretation. Research-grade WGS from sequencing providers costs considerably less but excludes clinical reporting, CLIA certification, and the variant interpretation that is the primary clinical value.

ConfigurationTypical US Clinical CostCPT / Notes
Proband WGS (standard)$2,000 – $4,000CPT 81425
Trio WGS (proband + both parents)$4,000 – $8,000CPT 81425 + 81426 ×2
Rapid / ultra-rapid WGS$5,000 – $10,000+Indication-dependent
Tumor-normal paired WGS (oncology)$3,000 – $6,000Indication-dependent

Insurance coverage for clinical WGS is less consistent than for WES and varies by payer, indication, and geography. Medicare coverage is limited and indication-specific; rapid WGS for critically ill neonates has the strongest coverage support, with several positive coverage decisions from MACs. Commercial coverage is expanding, driven by evidence from rapid WGS trials and ACMG guideline endorsement. Many commercial payers cover WGS for specific inpatient indications (critically ill neonates, NICU/PICU patients) with more restrictive criteria for outpatient rare disease. Prior authorization is required and medical necessity must specifically address why WGS is indicated over WES or targeted panel testing.

Common Questions

Frequently Asked Questions

What is the difference between NGS and WGS?
NGS (next generation sequencing) is the category of technology, referring to any massively parallel sequencing method that reads millions of DNA fragments simultaneously. WGS is one specific application of NGS that sequences the entire genome. Other NGS applications include WES, which sequences only protein-coding regions, and targeted gene panels. All WGS is NGS, but not all NGS is WGS. For the full pipeline view, see the NGS analysis guide.
Does sequencing test for Ehlers-Danlos syndrome?
Yes, and WGS is increasingly the preferred test for genetically complex EDS presentations. Several EDS subtypes have well-defined genetic causes: vascular EDS (COL3A1), kyphoscoliotic EDS (PLOD1, FKBP14), classical EDS (COL5A1, COL5A2). WGS detects all coding variants WES detects in these genes, plus deep intronic splice variants and regulatory variants that WES misses. Hypermobile EDS (hEDS), the most common subtype, has no identified genetic cause and cannot be diagnosed by any sequencing test. A clinical geneticist should guide test selection.
Why is nanopore sequencing different from Illumina for WGS?
Oxford Nanopore (ONT) and Illumina are fundamentally different sequencing technologies. Illumina reads short fragments (150 bp paired-end) with very high per-base accuracy (>99.9%) at lower cost per gigabase, the current standard for clinical WGS. Nanopore reads individual DNA molecules directly, producing reads of thousands to millions of base pairs without fragmentation or amplification. This enables superior detection of structural variants, repeat expansions, and complex genomic regions short reads cannot span, plus real-time direct DNA and RNA sequencing. Nanopore is not universally "better": for standard SNV and small indel detection, Illumina remains more accurate and more cost-effective. For SV characterization, repeat expansion analysis, and phasing, long-read platforms provide significant advantages.
Is Illumina or nanopore cheaper for WGS?
For standard clinical WGS, Illumina is currently cheaper on a per-gigabase and per-sample basis. The NovaSeq X produces 30x whole genome data at ~$200 to $400 per sample at scale. Oxford Nanopore and PacBio long-read WGS cost more per sample and produce less uniform coverage, making them more expensive for genome-scale clinical applications. However, when long-read WGS is the appropriate test (repeat expansion characterization, complex SV resolution, phasing), Illumina cannot substitute regardless of cost. The cost comparison is only relevant when both platforms can address the clinical question.
How long does whole genome sequencing take?
Standard clinical WGS has a turnaround time of approximately 3 to 8 weeks from sample receipt to report, varying by lab and configuration. Trio WGS may take slightly longer due to parental sample coordination. Rapid WGS programs serving critically ill neonates and pediatric ICU patients achieve 24 to 72 hour turnaround through purpose-built pipelines and dedicated sequencing capacity. Ultra-rapid programs have demonstrated results in under 14 hours in research settings.
What does whole genome sequencing find that exome sequencing misses?
WGS detects four major categories of variants WES cannot reliably identify: non-coding regulatory and intronic variants that affect gene expression or splicing; balanced structural variants (inversions, translocations) that require intact read-pair architecture to detect; repeat expansions at known disease loci; and variants in genomic regions WES capture probes cover poorly. About 15 to 25% of WES-negative patients with strong clinical suspicion of genetic disease who eventually receive a diagnosis do so through WGS detection of variants in one of these categories.
Can whole genome sequencing be used for cancer testing?
Yes. Tumor-normal paired WGS is one of the most comprehensive approaches to cancer genomic profiling, detecting somatic SNVs, indels, CNVs, structural variants (including gene fusions), mutational signatures, and tumor mutational burden in a single assay. It provides a complete somatic landscape that targeted panels cannot replicate. In clinical oncology, targeted panels remain more common for routine tumor profiling due to lower cost and faster turnaround, but WGS is used in research programs, comprehensive cancer centers, and for specific indications where panel testing has been non-diagnostic. For depth on somatic analysis see the somatic variant analysis guide.

Build a Clinical WGS Program

VarSeq handles the full complexity of clinical WGS tertiary analysis: multi-caller variant integration, non-coding annotation, ACMG-guided interpretation, and signed clinical reports. Designed for the scale, rigor, and turnaround demands of production clinical genomics.