Hypothesis to test: There is an excess of signal on the X chromosome compared to the autosomes in ADHD, autism and conduct disorder.

Methods to test hypothesis: First perform genome-wide association study on the number of ADHD symptoms for autosomes and X chromosome. Then two methods will be applied to test the hypothesis:

• Hypergeometric analysis based on p value thresholds to compare number of significant SNPs on X chromosome and autosomes. Since this analysis is not a fair comparison between the X chromosome and the autosomes (males have only half the alleles on the X compared to the autosomes and this means that X chromosome alleles would not reach the levels of significance observed for the autosomal alleles), another method will be used as well
• Simulations: Perform 1000 simulations of X (number of SNPs on the X chromosome) consecutive SNPs on the autosomes and calculate the number of significant SNPs for each p value threshold. Compare with number of SNPs on the X chromosome over same p value thresholds

Same analysis performed while also sampling half cases and half controls to make it more comparable with X where we have half the alleles for males.

The same methods will be used to test for an excess of signal on the X chromosome regarding autism and conduct disorder after performing relevant genome-wide association studies.