PurBayes

PurBayes estimates tumor purity and detects intratumor heterogeneity from next-generation sequencing (NGS) data of paired tumor–normal samples using a Bayesian finite mixture modeling framework with Markov Chain Monte Carlo (MCMC) inference.

Key Features:

• Bayesian Framework: Uses Bayesian inference with a Markov Chain Monte Carlo (MCMC)-based algorithm to quantify uncertainty in purity and heterogeneity estimates.
• Finite Mixture Modeling: Applies finite mixture models to deconvolve tumor and normal cell populations within sequencing data.
• Intratumor Heterogeneity Detection: Identifies subclonal variation within tumors to characterize intratumor heterogeneity and potential treatment-resistant clones.
• Paired Tumor–Normal Analysis: Analyzes paired tumor and normal samples to distinguish somatic alterations from germline variants and normal contamination.

Scientific Applications:

• Oncology research: Estimating tumor composition and intratumor heterogeneity to inform studies of cancer evolution and treatment resistance.
• Cancer genomics and precision medicine: Interpreting NGS-derived somatic variation across cancer types to support stratification and personalized treatment research.

Methodology:

Implements a Bayesian finite mixture modeling approach with MCMC-based inference and uses simulation-based validation of performance; the software is implemented as an R package.