PennCNV2

PennCNV2 detects copy number alterations (CNAs) in tumor samples using total and allele-specific signal intensity data from single nucleotide polymorphism (SNP) genotyping arrays to characterize aneuploidy, stromal contamination, genomic waves, and intra-tumor heterogeneity.

Key Features:

• Implementation: C++ software package for CNA detection from SNP genotyping array signal intensities.
• Tumor-specific challenges addressed: Accounts for aneuploidy, stromal contamination, genomic waves, and intra-tumor heterogeneity that complicate CNA detection.
• Stromal contamination estimation: Estimates stromal contamination via a maximum likelihood approach across multiple discrete genomic intervals.
• Signal intensity conditioning: Conditions on genome-wide signal intensity to account for aneuploidy and genomic waves.
• Hidden Markov model: Integrates total and allele-specific signal intensity at each SNP with physical maps to produce posterior inference of CNAs.
• Performance: Demonstrated improved accuracy and computational efficiency relative to existing methods on cancer cell lines and patient tumors.

Scientific Applications:

• Cancer genomics: Detection and characterization of CNAs in tumor samples from SNP genotyping arrays for studies of cancer susceptibility and tumor progression.
• Cell line and tumor sample analysis: Analysis and validation of CNA calls in cancer cell lines and patient tumor datasets.

Methodology:

Estimates stromal contamination using maximum likelihood across discrete genomic intervals, conditions on genome-wide total and allele-specific SNP signal intensities to address aneuploidy and genomic waves, and applies a hidden Markov model integrating signal intensities and physical maps for posterior CNA inference.