BACOM

BACOM estimates tumor-specific copy number alterations and normal tissue contamination from SNP array data using Bayesian analysis to recover cancer-cell copy number profiles for cancer genomics.

Key Features:

• Bayesian copy-number deconvolution: Performs Bayesian analysis of copy number mixtures to separate cancerous and normal cell contributions.
• Normal tissue contamination estimation: Estimates the fraction of normal cell contamination and corrects observed signals to recover true cancer-cell copy numbers.
• Genomic deletion detection: Detects genomic deletions within cancer genomes while accounting for normal cell admixture.
• Improved SCE detection: Increases power to detect significant consensus copy number events (SCEs) after in silico contamination correction.
• Data and validation: Analyzes SNP array data and has been validated on simulated datasets, prostate cancer datasets, and The Cancer Genome Atlas (TCGA) high-grade ovarian data.
• Implementation: Implements the pipeline as a cross-platform Java application with an R interface named bacomR.
• End-to-end pipeline: Provides end-to-end copy number analysis for heterogeneous cancer tissues, including relevant processing steps.

Scientific Applications:

• Tumor CNA analysis: Recovery and interpretation of tumor-specific copy number alterations from heterogeneous clinical samples.
• SCE discovery: Detection and improved statistical power for significant consensus copy number events (SCEs).
• Driver gene identification: Facilitates identification of potential oncogenes and tumor suppressors affected by CNAs.
• Cohort and TCGA analyses: Application to SNP array and TCGA datasets for cohort-level cancer genomics studies.

Methodology:

Bayesian analysis of copy number mixtures and in silico estimation and correction of normal tissue contamination from SNP array data, implemented as a Java pipeline with an R interface (bacomR).