Can you trust your reconstructed lineage tree? A homoplasy-based approach for irreversible evolution

Technical Summary: Can You Trust Your Reconstructed Lineage Tree? A Homoplasy-Based Approach for Irreversible Evolution

Problem Statement
The paper addresses a critical challenge in computational biology: assessing the reliability of inferred phylogenetic trees when the ground truth is unknown. While phylogeny inference is a well-studied problem with numerous algorithms, the emergence of single-cell genomics coupled with Cas9-based genome editing has created a new domain for cell lineage analysis. These techniques track processes ranging from embryogenesis to cancer by recording Cas9-induced mutations. A defining characteristic of this specific context is the non-modifiability of these mutations; once a Cas9-induced edit occurs, it cannot be reversed or overwritten. This irreversibility distinguishes cell lineage analysis from traditional phylogenetics and motivates a re-examination of how to validate tree reconstructions in the absence of known ground truth.

Methodology
The authors propose a homoplasy-based approach to evaluate the quality of reconstructed lineage trees. The method leverages the fundamental property that Cas9-induced mutations are irreversible. In this framework, homoplasy (the independent occurrence of the same mutation in different lineages) is treated as a signal of reconstruction error rather than a biological reality, given the unidirectional nature of the editing process.

The core of the methodology involves calculating a homoplasy score for a given tree. The authors posit that a tree with a lower homoplasy score is more likely to be accurate, as it implies fewer instances where the reconstruction forces independent mutations to explain the data—a scenario that contradicts the non-modifiable nature of the edits. The computation of this score is described as simple and scalable, making it feasible for large-scale datasets.

Key Contributions

1. Novel Validation Metric: The paper introduces a homoplasy-based score specifically tailored for irreversible mutation systems, offering a way to assess tree reliability without ground truth.
2. Superiority over Parsimony: The authors demonstrate that their homoplasy-based score is substantially more powerful than the commonly used parsimony score for distinguishing accurate reconstructions from incorrect ones in this specific context.
3. Theoretical and Empirical Validation: The claim of superiority is supported by both theoretical analysis and extensive simulations across a broad range of settings.
4. Scalability: The method is highlighted for its computational efficiency, enabling rigorous analysis of complex cell lineage data.

Results
Through simulations, the study shows that the proposed homoplasy-based method can effectively distinguish accurate lineage trees from a pool of candidate solutions. The results indicate that under diverse conditions, the homoplasy score outperforms the parsimony score in identifying the correct tree topology. The empirical data aligns with the theoretical analysis, confirming that the irreversibility of Cas9 edits makes homoplasy a highly informative metric for error detection.

Significance
The paper claims that this work opens the way for more rigorous analysis of cell lineages by providing a practical, scalable tool to validate phylogenetic reconstructions. By moving beyond traditional metrics like parsimony, the approach allows researchers to better trust the lineage trees generated from Cas9-based single-cell data. This is particularly significant for understanding complex biological processes such as clonal expansion, selection, and diversification, where the accuracy of the underlying tree is paramount. The authors maintain a focused scope, presenting the method as a robust solution for the specific constraints of irreversible mutation systems rather than a universal fix for all phylogenetic problems.