Genetic Heterogeneity in Autism Spectrum Disorder: Diagnostic Yield, Recurrent Genes, and Rare Variant-Phenotype Associations from Whole-Exome Sequencing

Abstract

Autism spectrum disorder (ASD) is genetically heterogeneous, involving rare and common variants that disrupt neurodevelopmental pathways. To explore this complexity, we performed whole-exome sequencing in children with ASD. Clinical phenotypes were systematically recorded, and severity was classified according to DSM-5 criteria. Variants were interpreted using ACMG guidelines, with recurrence analysis to identify genes shared across individuals and cohort enrichment testing against gnomAD. To examine genotype-phenotype relationships, we applied SKAT/SKAT-O across 16 phenotypes after covariate adjustment. Among 25 included individuals, pathogenic or likely pathogenic variants were found in 9, yielding a diagnostic yield of 36%. These involved genes linked to neurodevelopmental, epileptic, metabolic, and syndromic disorders. Recurrence analysis identified 586 genes present in at least two individuals, with PABPC1, GTF2I, PCLO, PKD1, and EP400 being the most frequent. SKAT/SKAT-O revealed the strongest burden associations for motor delay, aggressive behavior, mutism, anxiety, unresponsiveness to spoken voice, digestive disorder, and sleep disturbances, with limited overlap across phenotypes. Several recurrent genes also showed phenotype-specific associations. Overall, this integrative WES study provides clinically actionable diagnoses, highlights recurrent genes, and uncovers phenotype-specific signals, supporting convergent pathways with gene-level heterogeneity.