This workflow detects copy number variants and forks by question: germline CNVs -> GATK gCNV; somatic exome/panel -> CNVkit; WGS -> ASCAT/FACETS/PURPLE; cfDNA -> ichorCNA. It commits the build + target BED + assay-matched panel-of-normals, builds the reference before segmentation, fits purity/ploidy before integer calls, and diploid-centers before GISTIC2. CNVkit (below) is the somatic exome/panel path.
conda install -c bioconda cnvkit
# Or pip
pip install cnvkit
Tell your AI agent what you want to do: - "Detect CNVs from my exome sequencing data" - "Run CNVkit on my tumor-normal pair" - "Find copy number changes in my samples"
"Call CNVs using my pool of normals"
"Detect amplifications and deletions"
"Create a copy number profile for my sample"
"Plot the CNV scatter for chromosome 17"
"Create a heatmap of CNVs across samples"
"Generate a diagram showing CNV locations"
| Input | Format | Description |
|---|---|---|
| BAM files | Aligned reads | Tumor and/or normal |
| Target BED | BED file | Capture regions |
| Reference | FASTA | Genome reference |