#TAS#

Snakemake for Targeted Amplicon Sequencing analysis with CNVkit

Starting from .bam files obtained with IonTorrent technology, CNVkit is used to call CNV.

STRATEGY3_pool_normal_references_TAS

CNVkit batch strategy. Tumor samples vs pooled reference of normal samples.

input files

• BAM tumor : Mapped sequence tumor reads (.bam).
• BAM normal: Mapped sequence normal reads (.bam).
• Target regions : BED or interval file listing the targeted regions.
• Annotation file (.refFlat.txt)
• Fish file

output files

• CNV segment files (.cns)
• CNV segment files with integer copy numbers (.call.cns)
• MET_summary.tsv (comparison between gains and losses in NGS data vs FISH data for the gene of interest)
• CNVKIT_PLOTS (diagram, scatter, heatmap, density)

intermediate files

• antitarget empty file
• for each TUMOR sample the target coverage (.cnn)
• for each NORMAL sample the target coverage (.cnn)
• for each TUMOR sample the antitarget coverage (.cnn)
• normal reference : reference built from pooled normal samples (.cnn).
• copy number ratio files (.cnr)
• target genes with CN gain or loss (.gene.gainloss.txt)
• get number of genes in each cnv (.call.num_genes.cns)
• exploded gene table (call.genes.tsv)
• focus on gene of interest (.call.gene.tsv)
• ...

STRATEGY4_pairs_tumor_normal_TAS

CNVkit step by step strategy. Tumor samples vs pooled reference of normal samples.

input files

• BAM tumor : Mapped sequence tumor reads (.bam).
• BAM normal Mapped sequence normal reads (.bam).
• Target regions : BED or interval file listing the targeted regions.
• Annotation file (.refFlat.txt)

output files

• CNV segment files (.cns)
• CNV segment files with integer copy numbers (.call.cns)
• MET_summary.tsv (comparison between gains and losses in NGS data vs FISH data for the gene of interest)
• CNVKIT_PLOTS (diagram, scatter, heatmap, density)

intermediate files

• antitarget empty file
• for each TUMOR sample the taget coverage (.cnn)
• for each NORMAL sample the target coverage (.cnn)
• for each TUMOR sample the antitarget coverage (.cnn)
• normal reference : reference built for each normal sample (.cnn).
• copy number ratio files (.cnr)
• target genes with CN gain or loss (.gene.gainloss.txt)
• get number of genes in each cnv (.call.num_genes.cns)
• exploded gene table (call.genes.tsv)
• focus on gene of interest (.call.gene.tsv)
• ...

TEMPORARY FOLDERS

STRATEGY1.1

CNVkit batch strategy. Tumor samples vs pooled reference of normal samples. Feed the normal .bam files to batch. Slower than 1.2 and 2 in case of pooled reference of normals since the pooled reference is reconstructed from scratch for every tumor sample. Ok for tumor normal pair.

STRATEGY1.2

Mixed batch + step by step strategy. Feed a pooled reference to batch (.cnn) to batch. Faster than 1.1 in case of pooled reference of normals since the pooled reference is contructed only once.

STRATEGY2

CNVkit step by step strategy (no batch).
Tumor samples vs pooled reference of normal samples.