Phylogenetic Tree Inference: A Top-Down Approach to Track Tumor Evolution.

Area of Science:

• Computational Biology
• Cancer Genomics
• Bioinformatics

Background:

• Understanding cancer evolution requires analyzing tumor heterogeneity and progression using multiple biopsies.
• Existing phylogenetic inference tools often depend on accurate allele frequency data, which is challenging for clinical samples like FFPE.
• There is a need for robust methods that can infer cancer phylogeny even with limited or absent allele frequency information.

Purpose of the Study:

• To introduce Phylogenetic Tree Inference (PTI), a novel and user-friendly method for inferring tumor phylogenetic trees.
• To enable cancer phylogeny reconstruction using only the presence or absence of somatic mutations, bypassing the need for allele frequencies.
• To demonstrate the applicability of PTI across various datasets, including epigenetics, with binary feature-by-sample matrices.

Main Methods:

• PTI employs an iterative top-down approach to construct phylogenetic trees from multiple tumor biopsies.
• The method utilizes a binary matrix representing the presence (1) or absence (0) of somatic mutations across samples.
• It does not require allele frequency data, making it suitable for a wider range of clinical samples.

Main Results:

• PTI successfully infers phylogenetic tree structures from tumor biopsy data.
• Comparative analyses show PTI achieves performance comparable or superior to state-of-the-art methods like LICHeE, Treeomics, and BAMSE.
• PTI demonstrates a significantly shorter run time compared to existing methods.

Conclusions:

• Phylogenetic Tree Inference (PTI) provides an effective and accessible solution for reconstructing cancer evolutionary trajectories.
• Its ability to work without allele frequencies broadens the scope of phylogenetic analysis for clinical and research settings.
• PTI is a versatile tool applicable to diverse biological datasets with binary feature-by-sample structures.