How were reference levels selected for categorical variables with multiple levels?

[Shicheng-Guo]

Dear Prof. Neale,

Thank you so much for the great contribution to the community with Genebass database. Could you explain how the statistical models were set up in GeneBass to analyze categorical phenotypes in the UK Biobank whole exome sequencing study? Specifically: How were reference levels selected for categorical variables with multiple levels?

Thanks,

Shicheng