Fig. 3.

Power curve across R² values for PS1. Parameters set for all groups b₃ = g₁ = g₂ = $\sqrt{0.05}$, ra = 0.3; rc = 0.25; re = 0.3; rf = 0.25; Environmental variances were dependent on shared and additive genetic variances: e₁ was specified as $\sqrt{1-a_1^2-c_1^2}$ and e₂ as $\sqrt{1-a_2^2-c_2^2}$. Additive and shared variances for the groups: (A) cannabis use (a² 51%, c² 20%) (Verweij et al. 2010) and schizophrenia (a² 81%, c² 11%) (Sullivan et al. 2003); (B) BMI (a² 72%, c² 3%) (Rokholm et al. 2011) and major depression (a² 37%, c² 1%) (Scherrer et al. 2003); (C) alcohol use (a² 49%, c² 10%) (Verhulst et al. 2015) and heart disease (a² 22%, c² 0%) (Wu et al. 2014); (D) dyslipidemia (LDL) (a² 60%, c² 28%) (Zhang et al. 2010) and heart disease (a² 22%, c² 0%) (Wu et al. 2014). Vertical lines were added to represent R² for four PSs reported in recent papers: a, smoking (Pasman et al. 2022); b, BMI (Furlong and Klimentidis 2020); c, LDL (Kuchenbaecker et al. 2019); d, attention deficit hyperactivity disorder (ADHD) (Demontis et al. 2019)