Beyond opioid-related traits, we also identified strong genetic correlations with alcohol, nicotine, and cannabis-related traits, from initiation (nicotine, rg = 0.63, p = 7.14E−38; cannabis, rg = 0.35, p = 1.06E−06), to high levels of consumption (drinks per week, rg = 0.36; p = 2.11E−15), to misuse (AUDIT-P: rg = 0.42, p = 5.60E−08) and dependence (alcohol, rg = 0.74, p = 7.21E−08; the Fagerström Test for Nicotine Dependence, rg = 0.60, p = 2.98E−10; cannabis use disorder, rg = 0.63, p = 2.36E−13), and other smoking behaviors, such as maternal smoking around birth (rg = 0.64, p = 3.44E−26). Interestingly, hierarchical HCA with AGNES found that POU clustered closely with substance use disorders, as opposed to consumption phenotypes (Supplementary Fig. 3).
