Over the past two decades, cell biologists and computational biologists have worked to develop lineage tracing techniques that can decode the relationships between progenitor cells and their differentiated progeny. Despite steady progress, the field has long been constrained by the limited specificity and scalability of experimental labeling methods, as well as the insufficient accuracy and temporal resolution of computational analysis approaches. Transferring established lineage tracing strategies to complex biological systems and disease models remains difficult due to cell population heterogeneity, noise in high-dimensional omics data, and challenges in multi-omics integration. A comprehensive cell lineage tracing system—one that integrates high-performance experimental techniques with robust computational methods—is therefore needed to achieve high-precision labeling, high-dimensional data acquisition, and accurate reconstruction of lineage relationships.

A new review published in Quantitative Biology takes stock of where the field stands. The collaborative team—Shanjun Mao, Chenyang Zhang, Runjiu Chen, and Shan Tang from Hunan University, Xiaodan Fan from The Chinese University of Hong Kong, and Jie Hu from Xiamen University—systematically surveys both experimental and computational approaches to cell lineage tracing. On the experimental side, the review covers non-CRISPR-based methods and CRISPR–Cas9-based strategies; on the computational side, it examines non-velocity low-dimensional embedding methods and RNA-velocity-based approaches. For each category, the authors detail the technical principles, application scenarios, advantages, and inherent limitations. They also discuss the translational value of lineage tracing in developmental biology, regenerative medicine, and oncology, and outline core challenges and future directions—including the potential integration of artificial intelligence.

Figure 1 outlines a typical workflow for computational cell lineage tracing. The process begins with specific labeling of target cells using experimental techniques such as CRISPR–Cas9 genetic barcoding or fluorescent protein labeling, ensuring that markers are heritable and traceable throughout development or disease progression. Next, single-cell RNA sequencing or spatial transcriptomics is used to collect high-dimensional molecular profiles, capturing both gene expression and spatial localization of individual cells. Finally, trajectory inference tools—notably RNA velocity—are applied to analyze dynamic states and developmental trajectories, while lineage reconstruction algorithms build cell lineage trees, predict cell fates, and visualize cellular dynamics during development, regeneration, or tumorigenesis. The framework provides a roadmap for high-precision, large-scale investigation of lineage relationships and cell fate decisions, and may ultimately support translational applications in regenerative medicine and personalized cancer therapy.
DOI
10.1002/qub2.70006