Use of machine leaning approaches to explore genetic and phenotypic associations for Autism Spectrum Disorder

Introduction: Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder of high complexity ASD is characterized by impaired social interaction and communication and by stereotyped behaviors, and a high heterogeneity in clinical presentation. It is hypothesized that such complex heterogeneous phenotypic behaviors are associated with genetic factors. To further dissect the complex correlations between phenotype and genotype in ASD, in the current study we used powerful machine learning algorithms, like decision trees, to integrate clinical information from diagnostic instruments like the ADI-R and the ADOS as well as adaptive behavior and cognitive scales (VABS and WISC) with genetic data (Copy Number Variants, CNVs). ASD traits in parents were assessed using specific tools SRS and PSPQ.