Clonal Evolution and Mutational Signatures: Approaches, Challenges, and Perspectives for Cancer Progression Inference
Abstract
The rapid development of high-throughput sequencing has generated large volumes of cross-sectional cancer genomic data, providing an important basis for studying tumor evolution. Reconstructing cancer progression from these static observations, however, remains challenging because temporal information is not directly observed and because genomic and multi-omics measurements are affected by biological heterogeneity and technical noise. This review examines three linked components: cross-sectional cancer genomic data, mutational-signature-based cancer progression inference, and clonal-evolution-based cancer progression inference. We compare the biological assumptions, data requirements, inferential targets, sources of uncertainty, validation strategies, and clinical applicability of these approaches, and discuss how single-cell and spatial technologies can complement cross-sectional inference. Rather than treating inferred trajectories as direct observations of tumor history, we emphasize the conditions under which evolutionary information can be extracted from cross-sectional data and the limitations that remain. We further distinguish established methodological evidence from emerging clinical applications, with the aim of clarifying current capabilities and priorities for future inference of cancer progression.