CNApp

CNApp analyzes somatic copy number alterations (CNAs) in cancer by generating purity-corrected genome-wide CNA profiles and computing quantitative CNA scores to enable association with clinical and molecular variables.

Key Features:

• Purity-Corrected Genome-Wide Profiles: Generates genome-wide CNA profiles from purity-corrected segmented data derived from microarrays and next-generation sequencing.
• CNA Score Computation: Computes Broad Copy Number Alteration Scores (BCS), Focal Copy Number Alteration Scores (FCS), and Global Copy Number Alteration Scores (GCS).
• Machine Learning-Based Predictions: Leverages machine learning algorithms to classify samples from segmented CNA data and to predict clinical outcomes and molecular subtypes.
• Integration with Clinical and Molecular Data: Performs association analysis of CNAs with clinical features and molecularly annotated variables at sample and cohort levels.

Scientific Applications:

• TCGA pan-cancer analysis: Applied to The Cancer Genome Atlas (TCGA) pan-cancer dataset of 10,635 genomes to classify cancer types by tissue-of-origin using CNA patterns and to identify cancer-type-specific ranges of broad and focal CNA scores.
• Hepatocellular carcinoma: Reproduced recurrent CNAs reported in hepatocellular carcinoma.
• Colon cancer: Predicted colon cancer molecular subtypes and microsatellite instability (MSI) status using broad CNA scores and discrete genomic imbalances.

Methodology:

Uses purity-corrected segmented data from microarrays and next-generation sequencing to generate genome-wide profiles, computes BCS/FCS/GCS, applies machine learning algorithms for sample classification, and performs association analyses with clinical and molecular variables alongside statistical analyses.