The 2019 paper "CCR5-∆32 is deleterious in the homozygous state in humans" looked like it could use some replication, so I did some.
I used UK Biobank data, restricted to white British unrelated participants from a quality-controlled subset, N = 337,008.
The .do file is a stripped down version of the code I ran to produce the analysis and graph.
While I can't share the individual participant data, anyone with access to UK Biobank data (field IDs in teh .do file plus genetic data) should be able to run a replication.
The way we defined and excluded participants is explained in the University of Bristol quality control documentation for UK Biobank genetic data: https://data.bris.ac.uk/datasets/1ovaau5sxunp2cv8rcy88688v/UK%20Biobank%20Genetic%20Data_MRC%20IEU%20Quality%20Control%20version%202.pdf
##EDIT 2019-06-20##
I've updated the repo to include an updated analysis file, which does more analyses and includes another SNP.
I've also written a blog post about all this, and included a word document with the text.