IntroductionAn early evaluating system for autism spectrum disorder (ASD) severity is crucial. Questionnaire survey is challenging for accurately assessing the severity levels for ASD in children. MethodsOffspring with ASD-like phenotypes were induced by treating pregnant mice with Poly (I:C) at GD12.5 and the placentae corresponding to the offspring were obtained by caesarean. The autism severity composite score (ASCS) for offspring was calculated through behavioral tests. HE staining and immunohistochemistry were used to observe the morphology of placenta. Candidate biomarkers were identified by weighted protein co-expression network analysis (WPCNA) combined with machine learning and further validated by ELISA. Sperman's was used to analyze the correlation between biomarkers and metabolome. ResultsThe placental weight and mean vascular area of male offspring with ASD-like phenotypes were significantly decreased compared with typical mice. According to the WPCNA, four modules were identified and significantly correlated with ASCS of offspring. Two biomarkers (ASPG and DAD1) with high correlation with ASCS in offspring were identified. DiscussionVEGF pathway may contribute to sexual dimorphism in placental morphology within mice with ASD-like phenotypes in term. The placental ASPG and DAD1 levels could reflect ASD-like symptom severity levels in male/female mice offspring.