Population Reference Data in the Molecular Diagnosis of Pathogenic CAPN3 Variants: Contextual Evidence and...

Pathogenic variants in CAPN3 cause calpainopathy, but the diagnostic use of population-scale sequence data remains challenging, because allele frequency, inheritance assumptions, and variant classification do not directly predict clinical disease expression. This study evaluated how population reference data can inform interpretation of pathogenic CAPN3 variants in a molecular diagnostic context.

We analyzed CAPN3 variants in gnomAD v4.1.0, applied ClinVar-supported American College of Medical Genetics and Genomics (ACMG)/ Association for Molecular Pathology (AMP) categorization to identify pathogenic and likely pathogenic variants, mapped classified variants across annotated protein domains, and used cumulative allele frequency to contextualize expected disease burden under recessive and dominant inheritance models.

A total of 5185 CAPN3 variants were identified, of which, 399 were classified as pathogenic or likely pathogenic. These variants were distributed broadly across the gene and protein, and included missense, nonsense, and frameshift changes, supporting a heterogeneous diagnostic landscape rather than a single recurrent hotspot. Using a cumulative pathogenic/likely pathogenic allele frequency of 0.0059, the modeled disease burden was approximately 35 per 1,000,000 under an autosomal recessive framework and 17.5 per 1000 under a dominant framework.

The recessive estimate is more consistent with current clinical understanding of calpainopathy, whereas the dominant estimate likely overstates manifest disease, because penetrance, phase, and phenotype are not captured by aggregate reference-dataset variation. Population reference data can therefore strengthen molecular diagnostic interpretation of rare disease, but only when integrated with inheritance-aware, phenotype-informed clinical assessment.