As previously noted, we can estimate the theoretical de novo variant rate per individual (exome) by multiplying the observed rate per base pair by the total size of all RefSeq hg19 coding regions. By subtracting the theoretical rate, per exome, of de novo variants in controls from the theoretical rate in probands, we can then estimate the percentage of probands in whom a de novo variant is contributing to TD risk (Iossifov et al., 2014; Sanders et al., 2015). Based on this calculation, we estimate that 5.0% (95% CI 1.3%–8.7%) of cases have a de novo LGD variant and 11.6% (95% CI 2.4%–20.8%) of cases have a de novo damaging variant contributing to TD risk (Table 3). Similarly, 6.9% (95% CI 4.9%–8.9%) of ASD cases have a de novo LGD variant mediating ASD risk (Sanders et al., 2015).
