Whole-Genome Sequencing Outperforms Panel and Exome Tests for Diagnosing Myopathies

For thousands of patients living with unexplained muscle disease, the path to a diagnosis has long been a frustrating cycle of biopsies, inconclusive tests and repeated genetic panels that come back negative. A major French nationwide study now offers the strongest evidence yet that a single, comprehensive technology can break that cycle. In an analysis of 265 patients with genetically undiagnosed myopathies, researchers found that whole-genome sequencing delivered a conclusive molecular diagnosis for 26 percent of families who had previously eluded gene-panel and exome sequencing approaches — and, when strong candidate variants of uncertain significance were included, the effective diagnostic yield climbed to 44 percent. The findings, published in Genome Medicine as part of the French Genomic Medicine Initiative known as PFMG2025, suggest that genome sequencing should be considered a first-line or early diagnostic tool for neuromuscular disorders rather than a last resort.

Myopathies — diseases that primarily affect skeletal muscle — represent one of the most heterogeneous diagnostic challenges in clinical genetics. They can arise from defects in hundreds of different genes, spanning structural muscle proteins, ion channels, metabolic enzymes, mitochondrial machinery and RNA-processing factors. Clinical presentations overlap heavily: a patient with progressive weakness may ultimately turn out to have a dystrophinopathy, a metabolic myopathy, a congenital myopathy or an inflammatory condition with a genetic underpinning. This heterogeneity has long forced diagnostic laboratories to choose between narrowly targeted gene panels, which are cheap but blind to anything outside their curated gene list, and exome sequencing, which captures the protein-coding portion of the genome but systematically misses variants buried in introns, regulatory regions and structural rearrangements.

The study was conducted within PFMG2025, a national program piloted by the French Genomic Medicine Initiative to bring genome sequencing into routine clinical practice for rare diseases. Between July 2020 and October 2024, patients with myopathies who remained genetically undiagnosed after prior testing underwent genome sequencing at two designated clinical laboratories: SeqOIA in Paris and AURAGEN in Lyon. The retrospective analysis drew on clinical, paraclinical and genetic data collected across these platforms, making it one of the largest real-world assessments of genome sequencing for muscle disease conducted at a national scale. Crucially, the sequencing was performed as part of ordinary clinical care, not as a research protocol, which means the results reflect what health systems can realistically achieve rather than what is possible only under ideal laboratory conditions.

The headline result is deceptively simple: among 265 families, genome sequencing produced a conclusive molecular diagnosis — variants classified as likely pathogenic or pathogenic under American College of Medical Genetics and Genomics criteria — in 68 cases, a diagnostic yield of 26 percent. That figure might appear modest until it is placed in context. Every one of these patients had already passed through the existing diagnostic cascade, including, for many, gene panels and exome sequencing specifically designed to catch myopathy genes. A 26 percent yield in this heavily pre-tested population demonstrates that genome sequencing is not merely a marginal upgrade over existing methods; it reaches diagnostic territory that those methods structurally cannot access.

The technical reasons for this advantage lie in the fundamental differences between sequencing strategies. Gene panels and exome approaches rely on capture-based enrichment, which selectively amplifies exons — the roughly one to two percent of the genome that encodes proteins — before sequencing. Deep intronic regions, where pathogenic variants can disrupt splicing by creating cryptic exon insertion sites, are largely invisible to these methods unless specifically probed. Copy-number variants, complex rearrangements and variants in poorly captured exons are also systematically underdetected. Whole-genome sequencing, by contrast, reads across the entire genome uniformly, without capture bias, allowing laboratories to detect deep intronic splice-altering variants, exon-level deletions and duplications, and other non-coding pathogenic mechanisms in a single assay. This breadth proved decisive in the French cohort: the authors report that genome sequencing was necessary for the resolution of 20.5 percent of all cases with conclusive or candidate findings, because the responsible variants would not have been detectable using other established techniques.

Beyond the confirmed diagnoses, the study highlights the more nuanced contribution of variants of uncertain significance. When the researchers considered candidate VUS findings that nonetheless raised a strong diagnostic hypothesis — for example, variants in genes consistent with the patient’s clinical presentation, or variants whose pathogenicity is supported by emerging functional evidence but which do not yet meet formal ACMG criteria for pathogenicity — the effective diagnostic yield rose to 44 percent. In clinical practice, such findings are not endpoint diagnoses, but they can reshape patient management, guide surveillance for known complications, inform reproductive counseling, and direct families toward research studies and, increasingly, toward gene-specific clinical trials. As gene-specific therapies for neuromuscular diseases multiply, the difference between a confirmed diagnosis and a strong hypothesis is shrinking in practical importance, because both can determine eligibility for intervention.

The identification of deep intronic variants deserves particular emphasis, because these discoveries are rewriting the known architecture of neuromuscular disease. Pathogenic changes located thousands of base pairs away from the nearest exon were long considered unfindable at diagnostic scale, yet they are now recognized as substantial contributors to myopathies — most famously in dystrophinopathies, where deep intronic pseud exon activation accounts for a meaningful fraction of patients who previously tested negative on exons-focused assays. In the French cohort, the ability of genome sequencing to elucidate these complex variants, and to uncover previously undescribed pathogenic mechanisms, accounted for a large share of the diagnoses that panel and exome sequencing had missed. The authors argue that this capacity underscores the complex nature of neuromuscular genetic disorders and demonstrates how genome sequencing can deliver a genetic diagnosis within routine clinical practice rather than in specialized research settings.

The PFMG2025 program itself is a case study in how to operationalize genomic medicine at national scale. Funded by the French government through the National Research Agency under the Programme d’Investissements d’Avenir, the initiative connects a dense network of neuromuscular reference centers — spanning hospitals from Paris to Marseille, Lyon, Bordeaux, Grenoble and beyond — with centralized sequencing and analysis platforms. Clinicians across the FILNEMUS network prescribed genome sequencing for patients with suggestive phenotypes, encoded clinical information in standardized HPO terms, and returned findings through structured counseling pathways. This architecture allowed the study to capture the full diversity of myopathy subtypes and patient ages, from infants with congenital presentations to adults with slowly progressive weakness, and to measure performance in the messy reality of clinical medicine rather than in a curated cohort.

For patients and families, the practical consequences of a molecular diagnosis are difficult to overstate. A confirmed genetic cause ends the diagnostic odyssey — often a decade or more of investigations — and replaces uncertainty with a named condition that carries a defined natural history, evidence-based surveillance recommendations and, increasingly, targeted treatment options. Genetic diagnoses also ripple across families, enabling cascade testing of relatives, precise reproductive risk assessment and prenatal or preimplantation diagnostic options. In neuromuscular medicine specifically, the therapeutic landscape is changing rapidly: therapies for spinal muscular atrophy have transformed that disease’s trajectory, and gene-targeted approaches for Duchenne muscular dystrophy, limb-girdle muscular dystrophies and metabolic myopathies are advancing through trials. Every one of these interventions depends on knowing the exact gene involved — which makes the diagnostic yield of genome sequencing a direct determinant of who can access emerging treatments.

The French results add to a growing international consensus that the era of sequential, gene-limited testing is ending. Health systems that still route patients through panels first, then exomes, then genomes only after years of delay, are paying for repeated negative tests while withholding the one assay capable of detecting the full spectrum of pathogenic variation. The PFMG2025 findings — a 26 percent conclusive yield in a pre-tested myopathy population, rising to 44 percent with strong candidate findings, and more than one in five resolved cases achievable only through genome sequencing — make a data-driven argument for moving genome sequencing forward in the diagnostic pathway for muscle disease. As sequencing costs continue to fall and analysis pipelines for non-coding and structural variants mature, the question facing clinicians and payers is shifting from whether to adopt genome sequencing to how quickly the diagnostic cascade can be reorganized around it.

Subject of Research: Diagnostic yield of whole-genome sequencing compared with gene-panel and exome sequencing in genetically unresolved myopathies

Article Title: Genome sequencing improves diagnostic outcomes over panel and exome sequencings in myopathies: findings from the French PFMG2025 initiative

Article References: Verebi, C., Maino, A., Metay, C., Nectoux, J., Leturcq, F., Michel, L., Menassa, R., Pion, E., Cossee, M., Gorokhova, S., Krahn, M., Ramond, F., Fauré, J., Hersent, C., Malfatti, E., FILNEMUS PFMG2025 Consortium, Audic, F., Authier, F.-J., Barnerias, C., … Stojkovic, T. (2026). Genome sequencing improves diagnostic outcomes over panel and exome sequencings in myopathies: findings from the French PFMG2025 initiative. Genome Medicine. https://doi.org/10.1186/s13073-026-01769-w

Image Credits: AI Generated

DOI: 10.1186/s13073-026-01769-w

Keywords: genome sequencing, myopathy, diagnostic yield, PFMG2025, exome sequencing, gene panel, neuromuscular disorders, deep intronic variants, genomic medicine, France, rare disease diagnosis, variants of uncertain significance