Despite substantial recent progress in the discovery of multiple rare de novo and inherited causative mutations in over 100 genes, most of the genetic liability of autism spectrum disorder (ASD) remains unknown. Hilary Coon’s laboratory recently developed a novel mutation identification effort and successfully applied this to 519 whole genome sequencing (WGS) Simons Simplex Collection (SSC) families. Coon now proposes to target the missing heritability in ASD by extending this novel mutation identification effort to additional ASD cohorts.

Coon now proposes, in collaboration with the laboratory of Gabor Marth, to extend this methodology to the new 619-family SSC data set, creating a highly accurate ASD-specific de novo and inherited structural variant resource. Coon and colleagues will further stratify this data set by scoring each participant’s genome-wide background polygenic risk for >35 physical disease states, psychiatric conditions and behavioral attributes. This strategy will highlight individuals that lie at the extremes of polygenic distributions for one or more of the tested comorbidities/attributes, making it potentially easier to identify the novel causative variants that increase ASD risk on that specific background. Using a decision-tree approach, Coon’s group will then determine the relative predictive strength of WGS variants (aggregated into pathways), polygenic background and covariates in defining ASD cases within these genetic subgroups. This approach has been successful in other medical arenas (e.g., cancer and schizophrenia) for predicting genetic subsets for subsequent development of treatments targeted for specific subgroups. To further characterize WGS subgroups, Coon plans to replicate and refine these findings in other large ASD cohorts.

This multifaceted approach will create a valuable community resource that includes an accurate and exhaustive catalog of inherited and de novo structural variants from the SSC WGS data and proband/family subsets stratified by polygenic risk. By providing a multifactorial dissection of ASD risk, this approach has the potential to also uncover novel causative ASD variants, helping to improve our understanding, diagnosis and treatment of ASD.