PROJECT SUMMARY/ABSTRACT Autism spectrum disorder (ASD) frequently co-occurs with attention-deficit/hyperactivity disorder (ADHD). Individuals with ASD have a 22 times greater risk of having ADHD compared with those without ASD, and recent evidence suggests that ASD co-occurs with ADHD at a higher rate than with any other mental health disorder. The negative impact of this co-occurrence on the individual is substantial; those presenting with both disorders (ASD/+ADHD) show lower cognitive functioning, more severe social impairment, and greater delays in adaptive functioning than individuals presenting with ASD without ADHD (ASD/-ADHD). The overall rationale of this proposal is that a multidisciplinary integration of genomic, neuroimaging, behavioral, human stem cell, and machine learning approaches may reveal key insights into the mechanisms underlying the debilitating and common co-occurrence of ASD/+ADHD in children. The overall objective of the proposed work is to identify the etiological mechanisms underlying ASD/-ADHD and ASD/+ADHD. We hypothesize that children with ASD/+ADHD will have unique genetic, molecular, cellular, brain structural, and neurobehavioral features compared to children with ASD/-ADHD. This hypothesis will be tested through four specific aims: 1) to identify prospective longitudinal behavioral and neuroimaging predictors of ASD/+ADHD compared to ASD/-ADHD; 2) to characterize molecular and cellular features of neurons differentiated from induced pluripotent stem cells (iPSCs) generated from individuals with ASD/-ADHD and ASD/+ADHD; 3) to identify and quantify the overlapping genetic architectures for ASD and ADHD; and 4) to develop a machine learning model integrating multi-modal data to predict ASD/-ADHD and ASD/+ADHD. Innovations of the proposed study include the application of state-of-the-art neuroimaging (optimized to facilitate brain imaging in difficult-to-scan populations), a prospective longitudinal design (to account for individual differences in the developmental course of ADHD symptoms as children with ASD age), iPSCs (to identify distinct cellular and molecular profiles), novel statistical methods for multi-phenotype modeling and gene identification, and an innovative multiview machine learning approach that integrates multi-modal data to identify the functional genomic elements and gene regulatory networks that underlie the emergence of ASD/+ADHD. This project is highly responsive to the IDDRC RFA, as it involves comprehensive -omic approaches to markedly increase our understanding of more than a single IDD condition to improve diagnosis and to facilitate future biomarker development. The knowledge gained will be significant because it can be used to inform a far more powerful multi-modal assessment of ASD and ADHD that integrates behavioral observations with technically advanced (but highly feasible) biological assays. These findings will have important implications for early screening and diagnosis of ASD an...