Fig. 3: Bivariate genetic correlation estimates between pneumonia susceptibility and psychiatric phenotypes.

Genetic correlation was estimated between ten psychiatric GWAS and pneumonia susceptibility GWAS. The forest plot denotes the linkage disequilibrium genetic correlation estimate, whilst the error bars are the 95% confidence internal. The different pneumonia GWAS effect sizes utilised with the panel of psychiatric disorders was as follows: top left ā genome-wide meta-analysis of self-reported and clinically ascertained pneumonia; top right ā genome-wide meta-analysis of self-reported and clinically ascertained pneumonia genetically conditioned on smoking heaviness (cigarettes per day); bottom left ā clinically ascertained pneumonia from FinnGen, and bottom right ā self-reported pneumonia from 23andMe. The psychiatric phenotypes were as follows: TSā=āTouretteās syndrome (Nā=ā14,307), SZā=āschizophrenia (130,644), PTSDā=āpost-traumatic stress disorder (Nā=ā174659), OCDā=āobsessive compulsive disorder (Nā=ā9725), cognition = general cognitive ability (intelligence, Nā=ā269,867), BIPā=ābipolar disorder (Nā=ā51,890), ASDā=āautism spectrum disorder (Nā=ā36,350), ANā=āanorexia nervosa (Nā=ā72,517), and ADHDā=āattention/deficit-hyperactivity disorder (Nā=ā53,293).