Parsing the heterogeneity of brain-symptom associations in autism spectrum disorder via random forest with homogeneous canonical correlation

•There are ASD subtypes with distinct patterns of brain-symptom associations.•A random forest based model is proposed to construct subtypes based on the relationship between two sets of multi-variables.•Heterogeneity of ASD might reflect changes in different subsystems of the social brain.•Subtyping of ASD can improve the accuracy of case-control classification.