Glatt, Stephen; Tylee, Daniel (2017)
      Autism is a complex neurodevelopmental syndrome that can be challenging to identify in young children. Family-based and genetic studies indicate that autism has a strong genetic component, though immunologic processes may also contribute to altered neurodevelopment. Over the past two decades, genome-wide investigations have provided critical insights into the genetic causes and molecular correlates of autism at the levels of both the individual and the population. Studies of DNA have identified highly penetrant genetic factors that appear to explain a sizable minority (20-40%) of autism-affected individuals. However, patho-developmental mechanisms are less-clearly understood for the remaining majority of individuals for whom no highly penetrant factors are identifiable(i.e., idiopathic autism). The present body of work contains three studies that used genome-wide assessment methods to predict the diagnosis of autism and to characterize its molecular correlates using samples that predominantly reflect idiopathic etiology. In Chapter 1, we demonstrate that large numbers of low-penetrance genetic factors(i.e., commonly varying single nucleotide polymorphisms; SNPs) can be harnessed with machine-learning methods to predict autism risk. Throughout this document, we provide a review and critical evaluation of DNA-and RNA-based autism biomarkers. In Chapters 2 and 3, we use microarray and RNA sequencing to identify consistent patterns of autism-related gene expression in peripheral blood samples. These patterns shed light on altered immunologic signaling and also implicate signaling pathways that are known to be involved in neurodevelopment and to influence autism-related clinical and neurobiological phenotypes. Importantly, our findings indicate that the molecular correlates of idiopathic autism may converge with mechanisms understood from high-penetrance genetic causes. Furthermore, our findings support the idea that immunologic mechanisms could contribute to perturbations of neurodevelopmental pathways. We integrate our findings in the context of existing literature, highlight current gaps in knowledge, and propose future experiments to address these questions.