Table 2 Role of de novo CNVs in simplex and multiplex autism.

From: Rates of contributory de novo mutation in high and low-risk autism families

  

SSC unaffected

affected

Effect

Group

CNV number

CNV rate

CNV number

CNV rate

Expected CNVs number

Delta

p-value

AD

PC

all

SSC affected

86

0.046

157

0.084

85.8

71.2

2 × 10−06

3.81% (2.16–5.47)

45.4% (29.2–57.9)

 

AGRE affected

  

56

0.051

50.8

5.2

0.29

0.47% (−1.15–2.13)

9.3% (−29.3–34.4)

coding

SSC affected

44

0.023

106

0.057

43.9

62.1

4 × 10−07

3.32% (2.04–4.63)

58.6% (40.9–71.0)

 

AGRE affected

  

34

0.031

26.0

8.0

0.12

0.72% (−0.52–2.06)

23.6% (−21.3–52.1)

intergenic

SSC affected

26

0.014

34

0.018

25.9

8.1

0.15

0.43% (−0.34–1.23)

23.7% (−24.9–55.6)

 

AGRE affected

  

15

0.014

15.4

−0.4

0.52

−0.03% (−0.93–0.83)

−2.4% (−122.3–43.7)

genic noncoding

SSC affected

16

0.009

17

0.009

16.0

1.0

0.41

0.06% (−0.56–0.70)

6.1% (−92.9–57.3)

AGRE affected

  

7

0.006

9.5

−2.5

0.74

−0.22% (−0.80–0.43)

−35.0% (−316.3–43.3)

  1. The table presents our results for the contribution to autism of de novo CNVs as a whole (rows labeled ‘all’ in the “effect” column) and separately for the subsets of CNVs labeled as coding (CNVs that overlap with a coding exon), intergenic (CNVs that do not affect any genes) and genic noncoding (CNVs that affect genes but not coding regions). We analyzed only de novo CNVs of at least 4KB that we could detect with identical power in SSC and AGRE (see Supplementary Fig. 4). We used the number of children in a group as normalization factor to compute the “expected CNVs number” under the null model and we show the rate of the de novo CNVs in three groups in the “CNV rate” columns. Otherwise, we used the same three groups of children and the same statistics (delta, p-value, AD, PC) to quantify the role of a class of de novo mutation as described in Table 1’s legend.
Back to article page