Table 1.

Genes with significant excess of DNVs across phenotype and sex groups

Group	Samples	No. DNVs			FDR 5% significant genes (q < 0.05)					FWER 5% significant genes (P < 3.64e–7)
		dnLGD	dnMIS	dnMIS30	CH model	denovolyzeR	DeNovoWEST	Union	Intersection	CH model	denovolyzeR	DeNovoWEST	Union	Intersection
Offspring phenotype group
ASD	15,560	1,661	9,187	1,454	70	52	113	133	41	26	22	27	39	16
DD	31,052	4,931	20,588	4,084	352	293	414	511	233	181	166	196	241	136
NDD*	46,612	6,592	29,775	5,538	399	323	479	615 (LC615)	237 (MC237)	208	171	190	264	138 (HC138)
Sibling	5,241	329	2,999	408	3	6	—	7	—	1	1	—	2	—
Proband sex group
Male	29,704	3,820	18,590	3,343	233	191	308	385	141	117	105	104	149	77
Female	16,530	2,716	10,986	2,149	222	193	208	277	152	117	112	109	144	85

Two levels of significance, FDR 5% and FWER 5%, were assessed based on the union and intersection of three statistical models. The lower confidence set (LC615) is based on the union of all genes at FDR 5% significance observed in any one or more of the three models; the MC237 set is based on the intersection of FDR 5% supported by all three models; the highest confidence set (HC138) is based on the intersection of genes with FWER 5% significance supported by all three models. The FDR 5% significance threshold was corrected by the Benjamini-Hochberg method for all of the genes in each model (18,946 genes in CH model, 19,618 genes in denovolyzeR, and 18,762 genes in DeNovoWEST); the FWER 5% significance threshold was corrected by the Bonferroni method for 19,618 genes (the largest among three models) and seven tests in the analyses (tests for dnLGD, dnMIS, and dnMIS30 variants in CH model; dnLGD and dnMIS variants in denovolyzeR; and DNVs and dnMIS variants in DeNovoWEST). Genes with DNV significance in siblings were excluded from all sets (n = 7).

^*NDD = ASD + DD. Details regarding the NDD risk genes can be found in Datasets S2–S5, S9, and S10.