Artificial Intelligence-Powered Insights into Polyclonality and Tumor Evolution

Recent studies have revealed that polyclonality—where multiple distinct subclones cooperate during early tumor development—is a critical feature of tumor evolution, as demonstrated by Sadien et al. and Lu et al. in Nature (October 2024). These findings show that early polyclonal interactions can overcome fitness barriers, ultimately transitioning to monoclonality as dominant clones emerge. Understanding and targeting these interclonal dynamics offers new therapeutic opportunities. In this perspective, we outline how computational modeling and artificial intelligence (AI) tools can provide deeper insights into tumor polyclonality and identify actionable therapeutic strategies. By applying ligand–receptor interaction analysis, clonal trajectory reconstruction, network and pathway modeling, and spatial analysis, researchers can prioritize communication hubs, evolutionary bottlenecks, and microenvironmental niches that sustain tumor progression. These approaches, when integrated with experimental validation, offer a translational pathway from foundational discoveries to personalized cancer treatments aimed at disrupting cooperative subclonal ecosystems and preventing malignant progression.We commend the recent Nature publications, “Polyclonality overcomes fitness barriers in Apc-driven tumorigenesis” by Sadien etal. [1] and “Polyclonalto-monoclonal transition in colorectal precancerous evolution” by Lu et al. [2], both featured on 2024 October 30. These groundbreaking studies employed distinct lineage tracing methods to investigate the origins and evolutionary dynamics of colorectal and intestinal tumorigenesis. Despite their different approaches, both studies reached convergent conclusions: Polyclonality plays a pivotal role in the early stages of tumor development, providing critical insights into how diverse cellular populations collaborate to overcome fitness barriers and drive tumor progression.