Dutch population structure across space, time and GWAS design

We studied fine-grained population genetic structure and demographic change across the Netherlands using genome-wide single nucleotide polymorphism data (1,626 individuals) with associated geography (1,422 individuals). We applied ChromoPainter/fineSTRUCTURE, identifying 40 haplotypic clusters exhibiting strong north/south variation and fine-scale differentiation within provinces. Clustering is tied to country-wide ancestry gradients from neighbouring lands and to locally restricted gene flow across major Dutch rivers. Despite superexponential population growth, north-south structure is temporally stable, with west-east differentiation more transient, potentially influenced by migrations during the middle ages. Within Dutch and international data, GWAS incorporating fine-grained haplotypic covariates are less confounded than standard methods.