The overall strategy of the research proposed here is to jointly exploit three large repositories of neuroimaging and neurobehavioral data for the purpose of identifying and characterizing brain and behavioral biomarkers of autism spectrum disorders (ASD). Each of the data repositories to be accessed has been developed by one of the three project Principal Investigators. They are (1) the BrainMap database, developed by Peter Fox at the University of Texas Health Science Center at San Antonio (UTHSCSA); (2) the autism Center of Excellence (ACE) neuroimaging archive, developed by Eric Courchesne at the University of California at San Diego (UCSD); and (3) the Genetics of Brain Structure (GOBS) neuroimaging archive, developed by David Glahn via an ongoing collaboration between UTHSCSA, the Texas Biomedical Research Institute and Yale University. Each of these repositories is funded by multiple federal awards. The BrainMap database has been in ongoing development since 1988. It contains the standardized locations of statistically significant functional (>88,000 reported locations; > 11,000 experiments) and structural (>16,000 reported locations; >2,400 experiments) effects reported in the peer-reviewed, coordinate-based neuroimaging literature. In the proposed project, the BrainMap database will be mined and modeled for candidate biomarkers in the form of connected networks implicated in ASD. Candidate neurobehavioral effects/symptoms associated with specific pathways will be extracted from the behavioral meta-data. These activities form Aim 1 of this proposal. The ACE neuroimaging archive has been in ongoing development since 2006. At present, it contains functional and structural magnetic resonance imaging (MRI) neuroimaging data and behavioral assessments from >1,200 infants, children and adults with ASD; age/gender-matched neurotypical controls; and non-ASD contrast patients. This is the largest fMRI dataset on infants and children in the world. These primary neuroimaging data will be tested for ASD-specific abnormalities in resting-state connectivity in candidate pathways identified by meta-analyses of the BrainMap database. Behavioral deficits associated with these pathways will be sought by analysis of comprehensive subject-associated clinical assessment data. These activities form Aim 2 of this proposal. The GOBS neuroimaging archive has been in development since 2006. It contains neuroimaging data, quantitative neuropsychological data, pedigree structure, and genetic data from >1,500 Mexican-American subjects who are members of ~50 large extended families. The overall purpose of this resource is to assess genetic influences on brain structure and function, beginning with heritability and co-heritability (pleiotropy) of brain and behavioral phenotypes and progressing to identification of quantitative-trait loci and gene discovery. In the present project, the GOBS archive will be used to assess heritability and pleiotropy of the resting-state networks and behavioral phenotypes implicated in ASD through Aims 1 and 2. These activities form Aim 3 of this proposal. The general strategy of using BrainMap meta-analyses to identify connected neural systems in healthy individuals and in various disease states is well established, with more than 300 peer-reviewed publications using this resource. The applicability of meta-analysis-derived models to guide heritability and pleiotropy analyses of the GOBS archive is also well established, being the topic of collaborative publications between the Glahn and Fox laboratories. The use of meta-analytic models to inform neuroimaging analyses of ASD is less well established, being the subject of recent collaborations between the Courchesne and Fox laboratories.