SV consensus pipeline

Created to call consensus SVs from the SV calls of GDAN DLBCL groups.

Usage

1. Fix `config.yaml`

samples_file, metadata has to be generated
ref should be relinked to your local references

2. Install prerequisites

Highly recommended that you install snakemake using the best practices in https://snakemake.readthedocs.io/en/stable/getting_started/installation.html
Since I haven't set up a setup.py yet, you could just type the following after having installed snakemake.

pip install tabix numpy pandas pybedtools wgs_analysis

3. Fix global variables in `Snakefile`

You may be interested in fixing CHROMS, SOURCES, SAMPLES to your liking.

4. Run

Now the fun part! A snakemake dry-run is highly recommended.

bash run_snakemake.sh

BEDPE output explanation

The output BEDPE files created by this pipeline has the following columns:
- #chrom1: chromosome of breakpoints 1
- start1, end1: position of breakpoint 1, in 0-based semi-noninclusive range (i.e. if a 1-based UCSC breakpoint is chr1:1, the breakpoint coordinate in the BEDPE is chr1 0 1
- chrom2, start2, end2: coordinates of breakpoint 2
- name: sources of the SV separated by __
- score: number of the sources of the SV
- strand1, strand2: orientation of breakpoint 1 and 2, respectively
- type: type of the SV, in {'DEL', 'DUP', 'INV', 'TRA'}
- gene1, gene2: gene annotation of breakpoint 1 and 2

Gene annotation of a breakpoint

For DEL and DUP

For breakpoint 1, the overlapping or the closest downstream gene that is encompassed by the deletion/duplication (i.e. has overlapping coordinates with the breakpoint pair interval) is annotated
For breakpoint 2, the overlapping or the closest upstream gene that is encompassed by the deletion/duplication is annotated

For INV and TRA

If a breakpoint orientation is upstream (i.e. '+'), the overlapping or the closest upstream gene to the breakpoint is annotated
If a breakpoint orientation is downstream (i.e. '-'), the overlapping or the closest downstream gene to the breakpoint is annotated