Table 1.

Association results of 8 independent SNPs that are significantly associated with lifetime cannabis use

SNP rs	Chr	Gene	BP	A1	A2	Freq A1	N	β	SE	p-value	Direction
rs2875907	3p12.1	CADM2	85,518,580	A	G	0.352	181,675	0.070	0.009	9.38e-17	+++
rs1448602	3p12.1	CADM2	85,780,454	A	G	0.756	184,765	-0.062	0.010	6.55e-11	---
rs7651996	3p12.1	CADM2	85,057,349	T	G	0.477	184,765	0.049	0.008	2.37e-09	+++
rs10085617	7p22.2	SDK1	3,634,711	A	T	0.416	184,765	0.046	0.008	2.93e-08	+++
rs9773390	8q21.13	ZNF704	81,565,692	T	C	0.933	44,595	-0.171	0.029	5.66e-09	--?
rs9919557	11q23.2	NCAM1	112,877,408	T	C	0.614	180,428	-0.055	0.009	9.94e-11	---
rs10499	16p11.2	RABEP2, ATP2A1	28,915,527	A	G	0.651	179,767	0.053	0.009	1.13e-09	+++
rs17761723	17p13.3	SMG6	2,107,090	T	C	0.346	184,765	0.047	0.009	3.24e-08	+++

Independent hits were defined as R²<0.01, window size 250 kb. The p-value threshold was set at p<5e-08 (conventional genome-wide significant threshold; significance was tested two-sided). Chromosomal region (Chr), gene the SNP is located in or the nearest gene (within 500kb; Gene), base pairs location SNP on Hg19 (BP), allele 1 (A1), allele 2 (A2), frequency of allele 1 (Freq A1), number of individuals for which variant was included (N), beta coefficient of the effect allele A1 (β), standard error (SE).

Direction per sample: allele A1 increases (+) or decreases (−) liability for cannabis use, or sample did not contribute to this SNP (?). Order of samples: ICC, 23andMe, UK-Biobank. Independent SNPs were selected as SNPs with linkage disequilibrium R²<0.1 using a window size of 250 kB.

SNP rs9773390 was not present in UK-Biobank sample and its effect is rather isolated (see Figure 1b and 2); it might not represent a robust association.