Confounder-Adjusted Differentiation of Colorectal Cancer via Dynamic Propagation of Pathway Influence

Colorectal cancer (CRC) exhibits profound molecular heterogeneity between left-sided and right-sided tumors with distinct therapeutic responses that current static genomic analyses incompletely explain. We developed Dynamic Functional Influence Computation (DynaFIC), a computational framework modeling time-resolved signal propagation through biological networks to quantify functional influence beyond static expression levels. Using the GSE39582 dataset comprising 583 primary CRC samples, we performed confounder-adjusted differential expression analysis controlling for microsatellite instability status, <i>BRAF</i> mutations, Tumor Node Metastasis (TNM) stage, age, and sex, identifying 105 laterality-associated genes that underwent DynaFIC temporal network analysis. Right-sided tumors exhibited dramatically higher network connectivity density despite fewer nodes, creating distributed vulnerability patterns with <i>HOXC6</i> as the dominant regulator, achieving 200-fold influence through network amplification. Left-sided tumors showed compartmentalized, hierarchical organization with <i>PRAC1</i> as the primary regulator and predictable expression-influence scaling. Temporal clustering revealed distinct propagation kinetics: right-sided tumors demonstrated rapid signal saturation requiring early intervention, while left-sided tumors exhibited sustained propagation permitting sequential approaches. Stability Volatility Index analysis showed right-sided tumors maintain significantly higher systemic vulnerability. These findings establish anatomical location as a fundamental network organizational principle, suggesting that incorporating temporal dynamics into cancer analysis reveals therapeutically relevant differences for precision medicine applications in colorectal cancer.