The missing mitochondrial DNA dimension in modern GWAS

Genome-wide association studies (GWAS) have transformed complex-trait genetics, but their analytical architecture has been built predominantly around the nuclear genome. Human genetic variation, however, is distributed across two genomic compartments, the nuclear and mitochondrial genomes, yet mitochondrial DNA (mtDNA) remains comparatively underrepresented in mainstream association frameworks. This asymmetry is increasingly difficult to justify as population-scale studies now show that mtDNA sequence variation, heteroplasmy, haplogroups, and mtDNA copy number (mtDNA-CN) can be interrogated at biobank scale and can yield reproducible phenotype associations. We propose a genome-aware framework that treats nuclear-focused GWAS (nGWAS), mtDNA-focused analyses (mtGWAS), and integrated mito-nuclear analyses (mnGWAS) as complementary layers of association genetics. An operational mnGWAS should apply mtDNA-specific quality control, model biologically distinct mitochondrial features separately, account for tissue source, blood-cell composition, ancestry, and technical covariates, evaluate nuclear and mitochondrial main effects, and then test prioritized cross-genome interactions with appropriate multiple-testing control and independent replication. Recent UK Biobank, gnomAD, Million Veteran Program, and whole-genome sequencing studies demonstrate that the necessary data and methods now exist. The next step is therefore not simply to append mtDNA to existing GWAS pipelines, but to extend association genetics across both genomic compartments in a way that can determine when mitochondrial variation contributes additional biological, mechanistic, or predictive information to complex-trait genetics.