TVC Local Pipeline

Usage

$ ./oncompnet.sh <analysis-name> <input-bam-file> <nThreads> <input-hotspots-bed-file (optional)>

Required Tools

The following tools are required

• BCFTools - Normalize vcf and breake multiallelic variants
• Docker - Container which runs the main part of the analysis
• R version 3.6.2 - Variant annotation and report generation
  • VariantAnnotation
  • BSgenome.Hsapiens.UCSC.hg19
  • TxDb.Hsapiens.UCSC.hg19.knownGene
  • org.Hs.eg.db
  • knitr
  • stringr
  • xtable
  • pander
  • ggplot2
  • timeSeries
  • readr
  • dplyr
  • data.table
  • httr
  • jsonlite
  • tidyr
  • grid
  • kableExtra
• WeasyPrint - Convert HTML rmarkdown output to pdf report

Directory Format

Directory	Content
datasets	Contains the analyses subdirs. Each analysis dir can contain multiple BAM files belonging to each analysis
params	TVC parameters json file presets. Currently hard-coded in oncopmnet shell script
bed	PMNet custom BED file
hotspots	Hotspot BED files to be used with the hotspot option
ref_genome	Hg19 reference genome files
tvc-out	Contains the output files organized in the same way as in datasets dir
scripts	Scripts used in the analysis

Running an analysis

Dataset creation

Analysis usually begins from a BAM file which is downloaded from local ION torrent server. The BAM file needs to be copied inside the analysis sub-directory under the datasets dir. Any required analysis sub-dirs will need to be manually created before the analysis. If the BAM file belongs to an already existing analysis sub-dir then it just needs to be copied to the right place.

WARNING: If the BAM file input in an analysis sub-dir already exists the analysis will exit with error. Generally, there can be no two identically named BAM files under the same analysis. If this is the case:

• either a different name must be given to the BAM file, or
• the BAM file should be assigned to a different analysis

Hotspot parameter

The option to provide a hotspots BED file is provided. In this case the name of the hotspot file should be provided as the 5th positional arguement, which is optional.

WARNING: The hotspots file must be inside the hotspots directory.

Analysis steps

The pipeline follows the next steps to produce the final report:

1. Creation of a temporary container based on this image: https://hub.docker.com/r/sgsfak/tmap-tvc
2. Inside the container
   • Preparation of Hotspots VCF with tvcutils (if needed)
   • Mapping with TMAP
   • Aligned BAM sorting with samtools
   • Sorted BAM indexing with samtools
   • Alignment stats generation with samtools
   • Variant calling with TVC
3. Normalization of vcf and splitting of multi-allelic variants with BCFtools
4. Variant annotation with Bioconductor VariantAnnotation
5. Creation of HTML report with Rmarkdown
6. Convertion of HTML report to PDF format with Weasyprint

Example

Running an analysis will require the following detailed commands after having followed instructions described in Dataset Creation chapter:

$ cd /media/galadriel/fleming/oncopmnet/oncopmnet_pipeline
$ ./oncompnet.sh test_analysis PMN_S1-Q1DP 12 ocp_hotspot_blist_generated.bed>