Two explanations may contribute to these minor differences in fold enrichment estimation between GPA and LMM: First, GPA only used summary statistics while LMM used both phenotype and genotype data. Second, GPA and LMM employed different mathematical definitions of fold enrichment: GPA used the ratio between and , while LMM used the ratio between the proportion of the variance explained by SNPs in the CNS set and the proportion of the CNS set in entire genome. Furthermore, we evaluated the significance of enrichment of the CNS set by hypothesis testing. As shown in Table 1, enrichment of the CNS gene set was strong in BPD and SCZ, moderate in ASD and MDD, and nonsignificant in ADHD.
