pedcbio-cnv-prepare

Modify annotated CNV files for pedcBio upload

Module author: Run Jin (@runjin326)

Currently, annotated CNV files consensus_wgs_plus_cnvkit_wxs_autosomes.tsv.gz and consensus_wgs_plus_cnvkit_wxs_x_and_y.tsv.gz (and the combined file consensus_wgs_plus_cnvkit_wxs.tsv.gz) only contain samples and segments that have CNV calls. To capture total number of samples profiled in PedCBio, this analysis does the following:

1. for samples without CNV calls, status == neutral was added to the results for all available segments
2. for samples with CNV calls, for diploid segments that were not present in the annotation file, status == neutral was added to the results for those segments

When adding back neutral segments to X or Y segments, based on the germline_sex_estimate (or reported_gender when germline_sex_estimate is not available), copy number and tumor ploidy were added as followed:

1. For female, copy_number=2, ploidy=2 and status=neutral were assigned to X segments
2. For female, copy_number=0, ploidy=0 and status=neutral were assigned to Y segments
3. For male, copy_number=1, ploidy=1 and status=neutral were assigned to both X and Y segments

Additionally, in the notebook, there are several steps that output the number of segment entries for each sample to make sure the right amount of segments were rescued.

Usage:

Rscript -e "rmarkdown::render('pedcbio_cnv_prepare.Rmd', clean = TRUE)"

Input:

• ../../data/consensus_wgs_plus_cnvkit_wxs_autosomes.tsv.gz
• ../../data/consensus_wgs_plus_cnvkit_wxs_x_and_y.tsv.gz
• ../../data/cnv-cnvkit.seg.gz
• ../../data/cnv-consensus.seg.gz
• ../../data/histologies.tsv

Output:

• results/consensus_wgs_plus_cnvkit_wxs_autosomes.tsv.gz
• results/consensus_wgs_plus_cnvkit_wxs_x_and_y.tsv.gz
• results/consensus_wgs_plus_cnvkit_wxs

The output files are directly uploaded to S3 buckets for loading into PedCBio.