ABSTRACT Sporadic loss-of-function mutations, recurrent missense mutations, and copy number variants (CNVs) contribute significantly to the etiology of autism, but much of the genetic architecture has not yet been understood. Most known pathogenic CNVs are large and the significance of many of these gene mutations is still not known. The goal of this proposal is to significantly increase the yield of high-impact autism mutations by focusing on the discovery of understudied classes of rare variants from whole-genome (n = 35,000 samples) and whole-exome (n = 150,000 samples) sequence data being generated from autism families. This proposal focuses on ultra- rare, gene-disruptive mutations and leverages the additional sensitivity afforded by whole-genome shotgun sequencing data, novel CNV discovery methods, and transmission of ultra-rare inherited mutations to increase yield of pathogenic mutations. Our target will include the discovery and validation of smaller and more complex structural variants (including CNVs) and private gene-disruptive mutations not enriched in de novo mutation but preferentially transmitted to autism children. We will assess the utility of high-fidelity long-read sequencing to discover more complex forms of structural variation that have been missed by standard short-read sequencing by investigating 100 unsolved cases with a higher likelihood of genetic risk. In addition, we propose to select 10 genes with evidence of de novo mutation for further clinical evaluation, phenotypic variability, and comprehensive genetic characterization. This will include five genes where only de novo mutations have been observed compared to five genes where both de novo and inherited mutations have been documented in order to understand carrier phenotypes. This proposal specifically focuses on the application of novel genomic methods, recurrent mutations, and inheritance patterns to discover pathogenic variants in order to develop a more sophisticated model to explain the genetic architecture of autism. As part of this effort, we will quantify and compare the risk of different classes of mutation for autism and investigate transmission disequilibrium differences. The end product of this analysis will be the identification and characterization of new classes of highly penetrant genic mutations that contribute significantly to etiology of autism, providing targets for clinical diagnostics and future therapeutics.