THE ALLELIC ARCHITECTURE OF RARE VARIATION IN AUTISM AND OTHER NEURODEVELOPMENTAL CONDITIONS

• Published in: European neuropsychopharmacology 2024-10, Vol.87, p.16-17
• Main Authors: Fu, Jack, Satterstrom, F. Kyle, McWalter, Kirsty, Brand, Harrison, Kueffner, Robert, Cutler, David, Samocha, Kaitlin, Robinson, Elise, Buxbaum, Joseph, Devlin, Bernie, Roeder, Kathryn, Kruszka, Paul, Sanders, Stephan, Daly, Mark, Talkowski, Michael
• Format: Article
• Language: English
• ISSN: 0924-977X
• DOI: 10.1016/j.euroneuro.2024.08.046

The fields of autism and neurodevelopmental disorder (NDD) genetics are rapidly advancing. Catalyzed by the power of large cohorts and integration of all classes of de novo and inherited protein-coding variation, dozens of genes have emerged to harbor variants that confer high relative risk for autism, and hundreds of genes have been associated with NDDs more broadly. Through examination of protein-truncating variants (PTVs), predicted damaging missense variation, and copy number variants (CNVs), our prior analyses have begun to map the allelic diversity of perturbations within 72 autism-associated genes and 373 genes associated with NDDs, finding intriguing evidence of genes with significantly higher mutation rates and differences in the distribution of clinical phenotypes in autism compared to NDD (Fu et al., 2022; Satterstrom et al., 2020). Despite this progress, cohort sizes remain insufficient for disentangling the shared and distinct genetic architectures of autism, NDDs, and other neuropsychiatric conditions, as well as associating genes with more subtle impacts on neurodevelopment. To advance these boundaries, we present the largest to-date study of rare coding variants, consisting of 62,013 autistic individuals, including 38,088 probands and 9,567 unaffected siblings from complete trio and quartet families, respectively, and 23,925 additional autism cases without parental information contrasted against 26,931 controls. By aggregating across the Autism Sequencing Consortium (ASC), the Simons Simplex Collection (SSC), the Simons Foundation Powering Autism Research (SPARK), and individuals from a leading diagnostic laboratory (GeneDx), this dataset totals almost 200,000 individuals, nearly a three-fold increase over prior studies. When we stratified the clinically-referred GeneDx autistic probands by co-occurring DD/ID status, we found synonymous, missense, and PTV de novo mutation rates in autism probands without DD/ID from GeneDx that were nearly identical to individuals ascertained for a diagnosis of autism in the ASC, SSC, and SPARK research studies (0.296 vs 0.294, 0.767 vs 0.763, and 0.141 vs 0.145 respectively), while GeneDx autism probands with DD/ID exhibited mutation rates similar to those observed in previous research studies of DD. Further analyses of these data solidified previous observations of significant enrichment of de novo PTVs among autism probands of 3x compared to siblings among the genes most intolerant to PTVs in the human ge