Bioinformatic investigation of convergent proteins and cellular functions downstream of autism-associated genes

Abstract Autism spectrum disorder affects 1 in 31 people yet exhibits remarkable genetic heterogeneity with over 1000 genes identified in autistic populations producing convergent neurobiological phenotypes through not yet understood mechanisms. Here, we employed systems-level network analysis to test if diverse autism risk genes converge onto common cellular vulnerabilities across brain cell types. Using 1238 autism-associated genes from the Simons Foundation Autism Research Initiative (SFARI) database, we stratified genes by cell-type-specific expression patterns and constructed protein-protein interaction networks to map functional clusters and identify critical connector proteins. Through permutation-based statistical analysis, we identified convergent downstream proteins targets, representing molecular vulnerabilities where multiple autism-associated protein pathways converge for miscoordination of cellular function. Network analysis revealed cell-type-specific vulnerability signatures: SYN1 in inhibitory neurons, GRIN/DLG/CAMK2 proteins in excitatory neurons, GRIA2A/DLG4/GRIN2 in oligodendrocyte precursor cells, ACTL6B in oligodendrocytes, and ARHGAP/DOCK proteins in astrocytes. Remarkably, these computationally predicted targets have been independently validated in human genetics and autism research, demonstrating that network-based convergent vulnerability analysis successfully identifies critical nodes where diverse genetic insults disrupt coordinated cellular responses. This framework provides a molecular basis for understanding autism pathophysiology and identifies promising targets for intervention in the treatment of neurodevelopmental disorders.