TitanCNA

TitanCNA infers subclonal copy number alterations (CNA) and loss of heterozygosity (LOH) from whole-genome sequencing (WGS) data to characterize intra-tumoral population structure and clonal prevalence.

Key Features:

• Probabilistic Modeling: Uses a Hidden Markov Model (HMM)-based probabilistic model to segment genomes and predict regions of CNA and LOH in mixed cell populations.
• Cellular Prevalence Estimation: Estimates the cellular prevalence of clonal clusters by modelling mixtures of tumor and normal cell populations.
• Population Structure Inference: Infers tumor population structure and evolutionary dynamics from patterns of CNAs and LOH across the genome.
• Experimental Validation: Predictions have been validated using fluorescence in situ hybridization (FISH) and single-cell sequencing data.

Scientific Applications:

• Cancer Genomics Research: Analysis of subclonal CNAs and LOH to study tumor heterogeneity and evolutionary dynamics.
• Clinical Oncology: Characterization of subclonal genomic events to aid interpretation of tumor genomic landscapes relevant to precision oncology.

Methodology:

Processes whole-genome sequencing data using HMM-based segmentation to identify regions of CNA and LOH, simulates and analyzes idealized mixtures of clonal populations from genomically heterogeneous tumor sites, and infers cellular prevalences of subclonal events.