Exploring candidate molecular links between aspartame and pulmonary arterial hypertension through network toxicology and transcriptomics

Objective To identify candidate molecular overlaps between aspartame target profiles and pulmonary arterial hypertension (PAH)-associated transcriptomic signatures. Methods We integrated network toxicology, Weighted Gene Co-expression Network Analysis (WGCNA), and 12 machine learning algorithms to prioritize candidate genes. Molecular docking, single-cell RNA sequencing, and virtual CAPN1 knockout analysis explored predicted interactions, cellular expression patterns, and network perturbations. Results We identified 21 overlapping genes enriched in renin–angiotensin and calcium-related processes. Machine learning prioritized CAPN1, ERAP2, ITK, PTGS1, and KDR. The five-gene panel achieved an area under the receiver operating characteristic curve of 0.972 in lung tissue and 0.660 in external blood samples. Docking scores ranged from −8.0 to −5.7 kcal/mol. PAH samples showed lower hematopoietic-cell proportions and reduced CAPN1 expression within this population. Virtual CAPN1 knockout identified VNN2 and MLX as candidate network effectors. Conclusion These analyses prioritize candidate genes for investigating molecular links between aspartame-related target profiles and PAH. The absence of exposure records, metabolite measurements, and experimental validation limits the findings to hypothesis generation.