Bayesian inference for copy number intra-tumoral heterogeneity from single-cell RNA-sequencing data

Area of Science:

• Genomics
• Computational Biology
• Cancer Research

Background:

• Copy number alterations (CNAs) are key drivers of tumor evolution and heterogeneity.
• Understanding intra-tumoral clonal structure at single-cell resolution is vital for effective cancer therapy.
• Existing methods for clonal analysis often require manual parameter input and separate cell clustering from CNA detection.

Purpose of the Study:

• To develop an automated Bayesian model for simultaneous cell clustering and CNA profiling in tumors.
• To identify the number of distinct tumor clones and their associated CNA events.
• To leverage single-cell RNA sequencing (scRNA-seq) data for comprehensive intra-tumoral clonal analysis.

Main Methods:

• Developed a Bayesian model integrating gene expression and germline single-nucleotide polymorphism data.
• Implemented a Gibbs sampling algorithm for model inference.
• Created an R package named Chloris for method implementation and accessibility.

Main Results:

• The developed model accurately clusters single cells into distinct tumor clones.
• It simultaneously infers CNA profiles for each identified clone.
• Demonstrated superior performance compared to existing tools in both cell clustering and CNA identification accuracy.
• Successfully applied to human metastatic melanoma and anaplastic thyroid tumor datasets.

Conclusions:

• The Bayesian model provides an automated and accurate approach to dissecting tumor clonal architecture.
• It reveals functional gene expression differences linked to CNA profiles across tumor clones.
• This method advances the understanding of intra-tumoral heterogeneity and has implications for personalized cancer treatment strategies.