miRNA-32 and Hippo Signaling in Cancer Stemness: Integrated Insights from Machine Learning and Bioinformatics

The interplay between epithelial-mesenchymal dynamics and stem cell-like behavior in tumors continues to be an area of active investigation. This study leveraged machine learning and bioinformatics to systematically identify miRNAs targeting EMT transcription factors and to elucidate their roles in cancer stemness. Candidate miRNAs were predicted using TargetScan, miRDB, and DIANA-microT-CDS, followed by a two-tier comparative analysis to refine specificity and minimize false positives. Their relevance in cancer was assessed via literature review. Predicted target genes were analyzed using gene ontology, pathway enrichment, and protein-protein interaction network tools. Among the 1, 389 predicted miRNAs, miR-32 emerged as a prominent candidate influencing multiple EMT -related transcription factors such as Twist1, Snail/Snai1, Slug/Snai2, Zeb1, and Zeb2. Further analysis revealed its potential to regulate key elements within the Hippo signaling pathway, including LATS1, LATS2, STK4, MOB1A, and MOB1B. Enrichment analyses also linked miR-32 to pathways such as TNF signaling, chromatin remodeling by HATs, platelet activity, and intercellular communication, highlighting its broader involvement in sternness-related mechanisms. Collectively, these findings position miR-32 as a pivotal modulator of cancer stem cell evolution via EMT and Hippo signaling pathways. Experimental validation is needed to explore its potential in addressing therapy resistance.