Dissecting the dynamics of cell states is crucial for understanding various biological processes, such as tissue development and tumor drug responses. Recent single-cell lineage tracing (scLT) technologies provide effective ways to track single-cell lineages through heritable cellular barcodes, while simultaneously detecting the molecular states of cells by sequencing. However, the analysis of scLT data remains a significant challenge due to the diverse data features and complex cell dynamics.

To address these challenges, a research team led by Jin GU published their new research on 15 October 2025 in Frontiers of Computer Science co-published by Higher Education Press and Springer Nature.

The team develops scLT-kit, a user-friendly Python package for automated processing and analysis of scLT data. It includes methods for quantifying data features, characterizing cell dynamics, and dissecting the mechanisms underlying cell fate decisions.
In the scLT-statistics module of scLT-kit, considering the barcode off-target or missing effect during lineage tracing experiments, it first calculates the barcoding fraction at each time point, and the proportion of barcodes inherited from pre-timepoint or transferred to post-timepoint. Besides, the clone sizes are assessed by counting the number of cells in them.

In the scLT-analysis module of scLT-kit, it first compares the distribution of transcriptomic similarities within- and across-clones to evaluate the clonal heterogeneity and analyze temporal changes in cell states. Then, according to the lineage tracing information across two time points, the cell-cell lineage relationships and dominant cluster-level fates are inferred and represented. Next, to evaluate cell fate diversity across different datasets, four indicators are established to measure the randomness in cell fates, as well as consistency and similarity among neighboring cells. Finally, the differential molecular characteristics between the subclusters with distinct dominant fates within each cluster are identified to uncover biological mechanisms underlying cell fates.

To demonstrate its robustness and effectiveness, scLT-kit is applied to multiple real datasets, including hematopoietic progenitor cell differentiation, C. elegans embryogenesis, embryonic fibroblasts reprogramming, and lung cancer cell lines under Osimertinib or erlotinib treatment. Further, they systemically compare the dynamic characteristics in normal developmental processes with those occurring in response to external perturbations.

Further work can integrate novel algorithms, not only to overcome the inherent limitations of scLT data but also to achieve more systematic dissection of various biological processes.

DOI: 10.1007/s11704-025-41249-9