Impact of IQ on the diagnostic yield of chromosomal microarray in a community sample of adults with schizophrenia

Table 2 Six gene-sets showing enrichment for rare duplications in the schizophrenia-LIQ group compared to the schizophrenia-average IQ group

		Schizophrenia-LIQ (n = 162)^a			Schizophrenia-average IQ (n = 278)^a			Analyses
			Participants			Participants		Analyses
Gene-set name	Total genes (n)	CNVs (n)	n	%	CNVs (n)	n	%	p	BH-FDR	OR
GO Nervous system development	1874	44	35	18.5	29	28	10.1	0.013	0.0783	1.7
Union inclusive	2874	56	45	23.8	43	41	14.7	0.058	0.1554	1.4
NMDAR^b	62	4	4	2.1	0	0	0	0.078	0.1554	Inf
Targets of FMR1, Darnell et al. (2012)	840	25	24	12.7	21	20	7.2	0.371	0.5563	1.3
GO Synaptic	622	15	15	8.5	11	11	4.0	0.471	0.5650	1.2
GO Nervous system transmission	716	18	17	9.0	10	10	3.6	0.761	0.7605	1.1

^aAll rare (< 0.1%) autosomal CNVs > 10 kb were included in the analysis; see “Methods” for details. Sample sizes represent those individuals with one or more rare CNVs overlapping at least 1 bp of coding sequence, according to RefSeq annotations, in each group (162/192, or 84.3%, of the schizophrenia-LIQ group; 278/325, or 85.5%, of the schizophrenia-average IQ group)
^bThe CNVs comprised four pathogenic 16p11.2 duplications in four individuals, for which the contributing gene for this gene-set result was MAPK3
LIQ low IQ, GO gene ontology, CNV copy number variation, p statistical result when all rare autosomal CNVs are included; BH-FDR Benjamini–Hochberg false discovery rate, Inf infinity, NMDAR N-methyl-D-aspartate receptor components