The computation of genetic correlation in LD score regression uses GWAS results from European ancestry meta-analysis studies for each phenotype (references in Table 1). The GWAS for nicotine dependence included eight studies from a meta-analysis of FTND (Hancock et al., 2015): Environment and Genetics in Lung Cancer Etiology Study (N=3,006, dbGaP accession number phs000093.v2.p2) (Landi et al., 2009; Landi et al., 2008); Collaborative Genetic Study of Nicotine Dependence (COGEND, N=1935 recruited from wave 1 and N=292 from wave 2, dbGaP accession number phs000092.v1.p1) (Bierut et al., 2007); Chronic Obstructive Pulmonary Disease Gene Study (N=2,211, dbGaP accession number phs000765.v1.p2) (Regan et al., 2010); UW-TTURC (N=1,534, dbGaP accession number phs000404.v1.p1) (Baker et al., 2007); Study of Addiction: Genetics and Environment (excluding COGEND participants, N=843, dbGaP accession number phs000092.v1.p1) (Rice et al., 2012); GAIN (N=774, dbGaP accession number phs000021.v3.p2) (Manolio et al., 2007); nonGAIN (N=671, dbGaP accession number phs000167.v1.p1) (Manolio et al., 2007); and the Dental Caries Study (N=243, dbGaP accession number phs000095.v2.p1) (Shaffer et al., 2011). Published GWAS results for schizophrenia, and four Tobacco and Genetics (TAG) Consortium analyses of smoking-related behaviors were downloaded from the Psychiatric Genetics Consortium website (https://www.med.unc.edu/pgc/results-and-downloads).
