CoNAn-SNV

CoNAn-SNV integrates copy-number state information into SNV detection to improve identification of single nucleotide variants that overlap copy number alterations in cancer genomes.

Key Features:

• Integration of Copy Number Information: Incorporates discrete copy number state as input to account for somatic segmental CNAs, amplifications, and the resulting extended genotype space and skewed allelic distributions.
• Binomial Mixture Models: Employs a panel of Binomial mixture models to model allelic counts from sequencing data, with the number of mixture components informed by the copy number state at each locus.
• Next-Generation Sequencing Compatibility: Operates on next-generation sequencing data, including whole-genome shotgun reads, to detect SNVs overlapping CNAs.
• Enhanced Sensitivity and Specificity: Demonstrates increased sensitivity for detecting somatic non-synonymous mutations without a corresponding loss of specificity as assessed by ROC analysis.
• Targeting Genomically Unstable Tumors: Specifically addresses SNV detection challenges in tumors with copy number instability and segmental duplications or gains.

Scientific Applications:

• Lobular breast cancer: Applied to whole-genome shotgun data from a lobular breast cancer case, identifying 21 somatic non-synonymous mutations missed by callers insensitive to copy number variation.
• Lymphoma genomes: Applied to a lymphoma genome with a relatively stable karyotype, matching other callers overall while showing superior sensitivity in regions of copy number gain.
• Characterization of tumor mutational landscapes: Used to improve detection of SNVs overlapping CNAs for more complete profiling of genomically unstable cancer genomes.

Methodology:

Integrates discrete copy number state as input and models allelic counts using a panel of Binomial mixture models with mixture counts determined by copy number state at each locus; performance is evaluated using ROC analysis.