SPhyR: tumor phylogeny estimation from single-cell sequencing data under loss and error.

Area of Science:

• Oncology
• Computational Biology
• Genomics

Background:

• Cancer is marked by intra-tumor heterogeneity, arising from distinct cell populations with unique somatic mutations like single-nucleotide variants (SNVs) and copy-number aberrations (CNAs).
• Single-cell sequencing technology provides high-resolution insights into these diverse cell populations.
• Understanding the evolutionary dynamics driving intra-tumor heterogeneity requires robust phylogeny estimation methods.

Purpose of the Study:

• To develop and present a novel computational method, Single-cell Phylogeny Reconstruction (SPhyR), for estimating tumor phylogenies from single-cell sequencing data.
• To address the challenge of mutation loss due to copy-number aberrations in cancer evolution.

Main Methods:

• SPhyR utilizes the k-Dollo evolutionary model, which accounts for mutation gain and subsequent loss.
• The method is grounded in a combinatorial framework connecting solutions to constrained integer matrix completions and the cladistic multi-state perfect phylogeny problem.
• The approach was evaluated on simulated datasets and a real-world metastatic colorectal cancer case.

Main Results:

• SPhyR demonstrates superior performance compared to existing methods on simulated data.
• The method successfully reconstructed tumor phylogenies, providing insights into cancer evolution.
• Validation on a metastatic colorectal cancer dataset confirmed its efficacy.

Conclusions:

• SPhyR is an effective tool for reconstructing tumor phylogenies from single-cell sequencing data.
• The k-Dollo model incorporated in SPhyR accurately captures mutation dynamics, including loss events.
• This method advances the understanding of intra-tumor heterogeneity and cancer evolutionary mechanisms.