Objectives: To identify common genetic variants underlying ASD by conducting a GWAS in a large, more adequately powered sample by combining ASD cases and trios from several studies and consortia. In addition, we aimed to explore the potential for genetic overlap or pleiotropy by searching for significant common variation shared between autism and schizophrenia.
Methods: We conducted a GWAS analysis for ASD in a data set comprising 3338 ASD trios, 161 cases and 526 controls. In addition, an analysis of “polygene scores” suggested that the bulk of SNPs showing association to risk for schizophrenia also showed some, albeit modest, association to risk for ASD. Therefore we carried out analyses to test whether and which risk SNPs identified as significant by GWAS of schizophrenia samples (i.e., 9394 cases and 12462 controls assembled by the PGC) showed association for ASD.
Results: Even in this large ASD sample, the GWAS meta-analysis revealed no genome-wide significant loci for ASD. However, when we examined the top hits from the PGC schizophrenia GWAS in the autism data set, we found that the direction of effects was the same in 37 out of 53 independent SNPs (p=0.00274). Additional analyses suggest that SNPs in three regions are highly associated with risk for both schizophrenia and ASD. These SNPs fall in or near mir137, TCF4, and MAD1L1, which will be described in detail.
Conclusions: Although we were not able to identify significant common variants for ASD alone, this PGC combined meta-analysis identified significant genetic overlap and specific markers shared by ASD and schizophrenia.