Objective
Understanding the molecular mechanisms of autism spectrum disorder (ASD) and its psychiatric comorbidities, including bipolar disorder (BD), is pivotal for uncovering pathways that shape neurodevelopmental trajectories and clinical heterogeneity. We aimed to identify ASD-specific gene-expression signatures and disrupted biological processes in prefrontal cortex, contrasting them with those observed in BD.

Methods
We performed a comparative transcriptomic analysis of RNA-seq datasets from postmortem prefrontal cortex samples of individuals with ASD or BD and controls. Differential expression was assessed with DESeq2, including batch as a covariate in the BD model. Functional interpretation used Gene Ontology over-representation analysis, KEGG Gene Set Enrichment Analysis and gene-concept network visualization.

Results
ASD samples showed 45 differentially expressed genes (DEGs), mainly downregulated non-coding RNAs, particularly small nuclear RNAs and small nucleolar RNAs. Enrichment analysis indicated a convergent profile related to RNA processing, spliceosome assembly and spliceosomal activity. In contrast, BD showed 12 candidate DEGs, mostly upregulated protein-coding genes. BD enrichment involved metal ion response and detoxification, amine and peptide hormone responses, vascular regulation and hydrolase activity, with genes associated with neuroinflammation such as SERPINA3 and CHI3L1 contributing to this profile. No shared DEGs or enriched GO Biological Process terms were observed between ASD and BD.

Conclusion
These results support transcriptomic divergence in the prefrontal cortex, with ASD characterized by spliceosomal dysregulation, contrasting with metal ion response, vascular regulation and inflammation-associated signals in BD. Our findings provide a transcriptomic framework for future studies investigating disorder-specific molecular mechanisms and candidate signatures in ASD and BD.