Human miRNAs in cancer: statistical trends and cross kingdom approach

MicroRNAs (miRNAs) are small non-coding RNAs that regulate gene expression post-transcriptionally and are frequently dysregulated in cancer. While most studies focus on individual miRNAs, global patterns and their potential cross-kingdom similarities remain underexplored. This study aims to identify statistically stable human miRNAs in cancer, their key target genes, and analyze sequence complementarity with plant miRNAs to highlight patterns for future research. Experimentally validated human miRNA–gene interactions from miRTarBase were integrated with TCGA expression data across multiple cancers. Using a nonlinear threshold (critical threshold III), 115 underexpressed and 93 overexpressed miRNAs were identified as regulators of 200 genes with the strongest dysregulation. Further, 10,898 plant miRNAs from 127 species were computationally compared to these human miRNAs, and average complementarity scores were calculated to identify plant miRNAs most similar to under- or overexpressed human miRNAs. Statistical parameters such as membership ratios and experiment counts quantified miRNA expression stability. Subsets of human miRNAs exhibited consistent over- or underexpression across cancers, with concordant target gene expression patterns. Several plant miRNAs showed higher complementarity to underexpressed human miRNAs, suggesting reproducible cross-kingdom sequence similarity patterns. Differences in complementarity were modest but systematic, providing a computational framework for prioritizing candidate miRNAs for further study. This work establishes a computational approach integrating human miRNA–gene interactions, cancer expression data, and plant miRNA sequences. It identifies statistically stable miRNAs, key target genes, and cross-kingdom sequence similarities without implying functional or therapeutic activity. The framework can guide future experimental studies in miRNA regulation, comparative genomics, and molecular evolution.