A comparison between CARLIN and DNA Typewriter in CRISPR-mediated lineage tracing

CARLIN and DNA Typewriter are two major breakthroughs in CRISPR-based lineage tracing technology. It is essential to understand the potential and performance of these methods in lineage tracing, which provides important guidance on experimental design.

In this study, we systematically compare these two strategies using a unified stochastic simulation framework with known ground-truth lineages. By explicitly modeling CRISPR editing dynamics, barcode evolution, and cell division processes, the framework enables quantitative benchmarking of lineage reconstruction accuracy across diverse experimental parameter regimes. Both methods are evaluated using multiple accuracy metrics, including Robinson-Foulds accuracy and triplet accuracy, allowing a comprehensive assessment of lineage reconstruction performance under various editing probabilities, sampling depths, and lineage lengths.

DNA Typewriter consistently outperforms CARLIN in lineage reconstruction accuracy when sufficient numbers of recording targets are used, particularly in more cell divisions. Sequential and ordered recording in DNA Typewriter substantially reduces ambiguity in lineage inference compared to unordered CRISPR barcode editing. CARLIN's lineage-recording potential exhausts rapidly under continuous induction, limiting its effectiveness in long-term lineage tracing. Triplet accuracy provides a more permissive and informative metric than Robinson-Foulds accuracy, especially under partial sampling scenarios.