Ayon Pal, Ranav Pal, Vivek Roy, Madhumita G. Chaki, Bhaskar Roy
The 5′ mRNA leader regulates translation through its length and upstream open reading frames (uORFs), but its systematic remodeling in cancer remains unclear. We analyzed isoform-weighted 5′ leader features across The Cancer Genome Atlas, the Genotype-Tissue Expression transcriptomes and the Clinical Proteomic Tumor Analysis Consortium proteogenomic cohorts. Tumor-associated shifts frequently opposed normal tissue aging trajectories, with the clearest signals in lung adenocarcinoma and lung squamous cell carcinoma. Raw closed-uORF burden defined the sharpest paired-supported lung cores, whereas residualized closed-uORF burden provided the most stable length-independent signal. Colon adenocarcinoma reproduced the transcriptomic pattern but showed a prominent adjacent-normal field effect, with nontumor colon already shifted toward the tumor pattern relative to healthy colon. Upstream–initiation features did not improve proteome-wide prediction beyond mRNA abundance but showed modest, gene-selective associations within independently prioritized age-opposed lung gene sets. Recurrent proteogenomic anchors included SEMA4A, TXNL1, PCDH7, ATP2A2, FOXJ3, and PLK1 . These findings define an age-opposed, tissue-dependent pattern of 5′ leader remodeling and prioritize specific genes and isoforms for functional investigation.