Analyse the Effect of controls selection

[olivierlabayle]

Controls are currently set to be all unrelated UKB participants with with no record of critical care. That is we are comparing ICU patients with a specific infection to the "general" population regardless of whether they had the infection. Alternatively, we could restrict controls to have also had the infection.

Need to:

• Define a control inclusion rule
• Analyse the GWAS results