Table 1.

Enrichment of CNS Gene Sets for Association Signal

	Ncase	Nctrl	Significance Threshold
			0.01				0.001
			Nexp	Nobs	p	Padj	Nexp	Nobs	p	Padj
All	8,139	10,469	1.3	23	< 0.001	< 0.006	0.2	13	< 0.001	< 0.006
Deletion	3,164	4,234	1.4	38	< 0.001	< 0.006	0.2	25	< 0.001	< 0.006
Duplication	4,975	6,235	1.4	14	0.004	0.024	0.2	10	0.001	0.006
All (minus known loci)	7,649	10,028	1.3	10	0.015	0.03	0.1	4	0.005	0.01
Deletion (minus known loci)	2,963	4,140	1.4	11	0.008	0.048	0.1	2	0.024	0.14
Duplication (minus known loci)	4,856	6,165	1.4	6	0.038	0.23	0.1	3	0.006	0.036

The number of CNS gene sets with association p value surpassing a pre-defined threshold (p < 0.01 or 0.001) was compared to that seen in permuted data (1,000 permutations of CNV case-control status). Columns list the number of case and control CNVs contributing to each analysis (N_case and N_ctrl, respectively); the average number of gene sets with p value surpassing a given threshold in the permuted data, N_exp; the actual number of gene sets surpassing the same threshold in the unpermuted data, N_obs; the empirical probability of finding N_obs or more gene sets surpassing the p value threshold in the permuted data, p; and the Bonferroni-corrected probability, P_adj. Results are given for the combined analysis of all CNVs and for the analysis of deletions and duplications separately; these are presented first for the full dataset and then for the subset of CNVs that do not overlap well-supported schizophrenia loci.