The Contribution of Rare De Novo and Inherited Coding Variants in Neurodevelopmental Disorders

Abstract

High-throughput sequencing technologies allowed for studying rare (allele frequency <5%) genetic variation previously inaccessible through genotyping arrays used in genome-wide association studies. While each rare variant explains a negligible amount of heritability, they can potentially better identify trait-associated genes. In this dissertation, we identified rare de novo and inherited coding variation via exome sequencing to discover genes in complex neurodevelopmental traits. First, we found that ~1/3 previously identified de novo variants were present as standing variation in the Exome Aggregation Consortium’s cohort of 60,706 adults; these recurrent de novo variants in aggregate did not contribute to risk for neurodevelopmental disorders. We further used a loss-of-function (LoF)-intolerance metric, pLI, to identify a subset of LoF-intolerant genes that contained the observed signal of associated de novo protein truncating variants (PTVs) in neurodevelopmental disorders. LoF-intolerant genes also carried a modest excess of inherited PTVs; though the strongest de novo impacted genes contributed little to this, suggesting the excess of inherited risk resides lower-penetrant genes. Working with the Autism Sequencing consortium, we analyzed rare de novo and inherited variants from the largest exome sequencing study of autism spectrum disorders (ASDs) to date (35,584 samples) to discover 26 Bonferroni significant genes and upwards of 102 genes (FDR<0.1). Comparing the frequency of deleterious de novo variants in ascertained ASD and ascertained intellectual disability / developmental disorders (ID/DD) samples, half of the identified genes conferred more risk to ID/DD than ASD and these two groups of genes have different phenotypic outcomes. Finally, aggregating genetic and phenotypic data for ID/DD, ASD, and congenital heart disease (CHD) individuals, we evaluated the effect of severe ID/DD on both de novo variant frequencies and gene discovery in ASD and CHD. Within ascertained ID/DD, comorbid ASD and CHD does not affect either the de novo variant frequency or the number of significant genes, but the converse was not true: ID/DD increased both the de novo variant frequency and the number of Bonferroni significant genes discovered in ASD and CHD.