HetRCNA: A Novel Method to Identify Recurrent Copy Number Alternations from Heterogeneous Tumor Samples Based on...

Area of Science:

• Genomics
• Cancer Research
• Bioinformatics

Background:

• Discovering cancer-associated copy number aberrations (CNAs) is crucial for understanding tumor development.
• Existing methods struggle to identify subgroup-specific CNAs in heterogeneous cancer cohorts.
• Recurrent CNAs (RCNAs) shared by most tumors are identifiable, but subgroup-specific RCNAs remain challenging.

Purpose of the Study:

• To develop a novel integrated method, HetRCNA, for identifying statistically significant subgroup-specific RCNAs.
• To simultaneously detect subgroup-specific RCNAs and their related subgroup samples from heterogeneous cancer data.
• To address the limitations of existing approaches in detecting RCNAs within specific cancer subgroups.

Main Methods:

• Utilized a matrix decomposition framework with weight constraints.
• Employed coefficients of left vectors (with weight constraint) to measure subgroup samples.
• Used coefficients of right vectors and significance testing to identify subgroup-specific RCNAs.

Main Results:

• HetRCNA demonstrated superior performance on simulated datasets compared to competing methods.
• The method proved robust against noise factors in simulated data.
• Application to a real breast cancer dataset successfully identified numerous RCNA regions, correlating well with existing approaches and highlighting cancer-related genes.

Conclusions:

• HetRCNA is an effective tool for identifying subgroup-specific RCNAs and associated samples in heterogeneous cancer.
• The method offers improved performance and robustness over existing CNA detection techniques.
• Identified genomic regions by HetRCNA in breast cancer samples contain known cancer-related genes, validating its clinical relevance.