Phylogenetic tree inference from single-cell RNA sequencing data with SCITE-RNA.

Area of Science:

• Computational Biology
• Genomics
• Evolutionary Biology

Background:

• Phylogenetic tree inference is crucial for understanding cellular evolution.
• Existing methods struggle with the complexity and noise of single-cell RNA sequencing (scRNA-seq) data.
• Accurate reconstruction of cell lineages is essential for linking genotype to phenotype.

Purpose of the Study:

• To introduce SCITE-RNA, a novel phylogenetic tree inference method tailored for scRNA-seq data.
• To improve the accuracy and robustness of phylogenetic analysis in single-cell studies.
• To enable the integration of evolutionary trajectories with gene expression profiles.

Main Methods:

• SCITE-RNA utilizes reference and alternative read counts of single-nucleotide variants.
• The method employs a maximum-likelihood random-scan greedy search algorithm.
• It alternates between cell lineage and mutation tree representations to avoid local optima and ensure convergence.

Main Results:

• SCITE-RNA demonstrated superior performance on simulated scRNA-seq data compared to existing phylogenetic inference methods.
• The method successfully reconstructed evolutionary trajectories in simulated datasets.
• Application to real cancer scRNA-seq data successfully linked cell evolutionary paths to gene expression patterns.

Conclusions:

• SCITE-RNA offers a significant advancement in phylogenetic tree inference for scRNA-seq data.
• The method provides a robust approach to uncovering cellular evolution and its relationship with molecular phenotypes.
• SCITE-RNA has practical implications for cancer research and understanding disease progression.