PICTograph

PICTograph models tumor clonal composition and infers subclone evolutionary relationships from multi-region sequencing data using a Bayesian hierarchical framework to estimate posterior distributions of cancer cell fractions (CCFs).

Key Features:

• Bayesian Hierarchical Modeling: Employs a Bayesian hierarchical framework that models uncertainty in assigning mutations to specific subclones and generates posterior distributions.
• Accurate Estimation of Cancer Cell Fractions (CCFs): Integrates multi-region sequencing data to produce consistent and precise posterior estimates of CCFs.
• Improved Clonal Tree Inference: Infers probable ancestral relationships between subclones and improves tree inference across varying levels of simulated clonal diversity.
• Visualization of Evolutionary Relationships: Uses ensemble-based visualization techniques to highlight highly probable evolutionary pathways among subclones.

Scientific Applications:

• Solid tumor evolutionary studies: Investigating intra-tumor heterogeneity and evolutionary trajectories in solid tumors to inform treatment strategies and prognostic assessments.
• Multi-region whole-exome sequencing of pancreatic lesions: Applied to multi-region whole-exome sequencing of pancreatic cancer precursor lesions, revealing 6–12 distinct subclones and evidence of intra-sample mixing.

Methodology:

Integrates multi-region sequencing data within a Bayesian hierarchical model to estimate posterior distributions of CCFs and reconstruct probable evolutionary trees of subclones.