Figure 2. The effect of PRS-IQ deciles on ASD risk and cognitive ability.

The estimate and 95% CI of each PRS-IQ decile on ASD risk and cognitive ability. Each regression accounted for the individual-level burden of deletions $\left(\sum \mathrm{D}\mathrm{E}{\mathrm{L}}_{\mathrm{p}\mathrm{L}\mathrm{I}}\right)$, duplications $\left.(\sum {\mathrm{D}\mathrm{U}\mathrm{P}}_{\mathrm{p}\mathrm{L}\mathrm{I}}\right)$, and – when available – cognitive ability of individuals included in the model. a) Even after adjusting for the effects of cognitive ability, a high PRS-IQ (10th decile) increases the risk for ASD, while a low PRS-IQ (1st, 2nd deciles) decreases the risk for ASD. b) A PRS-IQ below and above the 6th decile significantly decreases and increases cognitive ability in the general population, respectively. The effect of PRS-IQ on cognitive ability is the same in cases with ASD and in the general population (extrafamilial controls). Filled-in points represent statistically significant terms (P value ≤ 0.05 following FDR adjustment for multiple corrections). Error bars represent the 95% CIs. For detailed model results, see Table ST3. See Figure S6 and Table ST3 for the results across other analysis groups.