RT Journal Article SR Electronic T1 Contribution of de novo non-coding mutations to autism and identification of risk genes from whole-genome sequencing of affected families JF bioRxiv FD Cold Spring Harbor Laboratory SP 077578 DO 10.1101/077578 A1 Yuwen Liu A1 Ercument Cicek A1 Yanyu Liang A1 Jinchen Li A1 Rebecca Muhle A1 Nicholas Knoblauch A1 Martina Krenzer A1 Yue Mei A1 Yan Wang A1 Yi Jiang A1 Evan Geller A1 Zhongshan Li A1 Iuliana Ionita-Laza A1 Jinyu Wu A1 Kun Xia A1 James Noonan A1 Zhong Sheng Sun A1 Xin He YR 2016 UL http://biorxiv.org/content/early/2016/09/26/077578.abstract AB Analyzing de novo mutations (DNMs) in protein-coding genes from whole-exome sequencing (WES) data has emerged as a powerful tool for mapping risk genes of autism spectrum disorder (ASD). The impact of non-coding mutations in ASD, however, has been largely unknown. This represents a large gap in our understanding of the genetics of ASD, as the majority of GWAS hits for a range of disorders fall into non-coding regions. To address this question, we performed a meta-analysis of DNMs using whole-genome sequencing (WGS) data from more than 300 individuals with ASD. We found that DNMs are enriched within brain transcriptional regulatory elements near genes involved in neuropsychiatric disorders. In these genes and in evolutionarily constrained genes, we also found an excess of DNMs that are predicted to affect pre-mRNA splicing. Collectively, we estimate that non-coding mutations explain at least one third of the ASD genetic risk attributable to DNMs. By combining information of non-coding DNMs with published WES data, we identified three new ASD risk genes at a false discovery rate (FDR) < 0.1, and 11 at a FDR < 0.3. A number of these genes are known to regulate critical processes in neural development and have been associated with other neuropsychiatric disorders. Taken together, our results demonstrate the pathogenic contribution of non-coding DNMs in ASD etiology and highlight some promising ASD risk genes. The analytic tools we provided in this study, for estimating contribution of non-coding mutations to disease risk and for mapping susceptibility genes using both coding and regulatory mutations, are applicable to any WGS studies on DNMs.