This repo contains a snakemake pipeline to create a phased reference panel from 1kGP for imputation with GLIMPSE2. This pipeline follows their tutorial and use the most recent high coverage integrated phased panel from 1kGP.
Software requirements are detailed in the conda environment file (spec/snakemake.yml). Paths to the base directory where everything will be downloaded/written and to GLIMPSE should be set by editing spec/file_spec.json (by default, this will be written to a subdirectory outputs where this repo is downloaded). GLIMPSE uses genetic maps to help decide how to chunk the genome to. The install path for GLIMPSE and parallel must be provided in the spec file.
Note that the X chromosome is bugged, and I have not yet figured out how to fix it. For now, I am just excluding it entirely (both non-PAR and the two PARs).