The effects of three batch variables on marker genotypes were tested in iPSYCH2012 (ArrayPlate.ID, PreProc.Plate and wave) and iPSYCH2015i separately (Array.Batch, ArrayPlate.ID and PreProc.Plate). This was done using relatedness-pruned dataset with ancestry outliers removed to avoid removal of markers where batch effects were caused by population structure or cryptic relatedness rather than genotype artefacts.
