Table 1.

Role of de novo LGDs in simplex and multiplex autism.

	SSC unaffected		Affected
Group	Synonymous number	LGD number	Synonymous number	LGD number	Expected LGD number	Delta	p-value	AD	PC
SSC affected	484	157	499	283	161.9	121.1	8 × 10−07	6.48% (3.85–8.80)	42.8% (27.8–54.3)
AGRE affected			309	116	100.2	15.8	0.14	1.42% (−1.25–4.15)	13.6% (−13.4–35.1)

The table shows the numbers of de novo LGDs (“LGD number” columns) identified in 1,874 unaffected children from SSC, 1,869 affected children from SSC and 1,107 affected children the AGRE collection together with the number of de novo synonymous mutations (“synonymous number” columns), used for normalization, identified in the same children. The table also shows the expected number of de novo LGDs (“expected LGD number” column) in the two affected groups under the null model that the incidence of the such variants in the affected children is the same as the incidence in the unaffected children. The rest of the columns quantify the causal role of the de novo LGDs in the affected children from the simplex SSC and the multiplex AGRE collections. The “delta” column shows the excess of the de novo LGDs in the two affected groups. The “p-value” is the probability of achieving a delta larger or equal to the observed one under the null model, computed through permutation of the affect status. The “AD” column shows the ascertainment differential or our estimate of the percent of the affected children that have autism in part due to a de novo LGD; the “PC” column shows the percent contributory statistic that is the percent of the de novo LGDs identified in the affected children that contributed to the disorder. Finally, the table shows the 95% confidence interval for the AD and PC statistics computed by bootstrap. See the Results section for detailed description of the methods used.