Single-cell mutation calling and phylogenetic tree reconstruction with loss and recurrence

Area of Science:

• Genomics
• Computational Biology
• Cancer Research

Background:

• Tumours exhibit intra-tumour heterogeneity due to genomic aberrations, impacting relapse and treatment failure.
• Understanding tumour composition and evolution is crucial for developing personalized cancer therapies.
• Single-cell sequencing (SCS) offers high resolution for studying tumour heterogeneity but faces challenges with noise and evolutionary complexity.

Purpose of the Study:

• To develop a computational method for analyzing tumour heterogeneity using single-cell sequencing data.
• To accurately reconstruct tumour cell phylogeny and identify mutations, including losses and recurrences.
• To address the challenges posed by sequencing noise and complex evolutionary processes in SCS data.

Main Methods:

• A Bayesian approach is employed to model noise processes and account for mutation loss and recurrence during tumour evolution.
• The method jointly calls mutations in individual cells and reconstructs phylogenetic relationships.
• The approach quantifies certainty in mutation predictions.

Main Results:

• The developed method accurately calls mutations and quantifies prediction certainty.
• It successfully reconstructs phylogenetic relationships between cells.
• Demonstrated advantages in handling simulated data and applied to real tumour SCS data, showing improved analysis of tumour evolution.

Conclusions:

• The new method enhances the analysis of tumour heterogeneity from SCS data by modeling evolutionary dynamics.
• Accurate reconstruction of tumour cell phylogeny and mutation profiles is achievable.
• This approach facilitates a deeper understanding of tumour evolution for personalized medicine.