linear-mixed-model
JAMMA (Highly-Accelerated Multi-method Mixed-model Association) -- a fast and modern Python and C reimplementation of GEMMA for large-scale GWAS.