Skip to content

Latest commit

 

History

History
59 lines (45 loc) · 2.19 KB

README.md

File metadata and controls

59 lines (45 loc) · 2.19 KB
Raw

Nanopore-SV-Analysis

Structural variant filtering and analysis of Nanopore human WGS data.

nanopore-SV-analysis

Installation instructions

Download the latest code from GitHub:

git clone https://github.com/mike-molnar/nanopore-SV-analysis.git

Dependencies

There are many dependencies so it is best to create a new Conda environment using the provided YAML file:

conda env create -n nanopore-SV-analysis -f nanopore-SV-analysis.yml
conda activate nanopore-SV-analysis

Below is a list of the Conda dependencies:

  • BCFtools
  • bedtools
  • karyoploteR
  • cuteSV
  • Longshot
  • NanoFilt v2.8.0
  • NanoPlot v1.20.0
  • pybedtools
  • PySAM
  • PyVCF
  • seaborn v0.10.0
  • Snakemake
  • Sniffles2
  • SURVIVOR
  • SVIM
  • WhatsHap
  • Winnowmap2

Reference genome

You will need to download the reference genome manually before running the workflow. I have not included the download as part of the workflow because it is designed to run on a cluster that may not have internet access. You can use a local copy of GRCh38 if you have one, but the reference can only contain the autosomes and sex chromosomes, and the chromosomes must be named chr1, chr2, ...,chrX, chrY. To download the reference genome and index it, change to the reference directory of the workflow and run the download_reference.sh script:

cd /path/to/nanopore-SV-analysis/reference
chmod u+x download_reference.sh
./download_reference.sh

To run on a grid engine

Copy the Snakefile and config.yaml files to the directory that you want to run the workflow. Modify the information in config.yaml for your sample names and FASTQ locations. There are a few different grid engines, so the exact format to run the workflow may be different for your particular grid engine:

snakemake --jobs 500 --rerun-incomplete --keep-going --latency-wait 60 --cluster "qsub -cwd -V -o snakemake.output.log -e snakemake.error.log -q queue_name -P project_name -pe smp {threads} -l h_vmem={params.memory_per_thread} -l h_rt={params.run_time} -b y"

You will have to replace queue_name and project_name with the necessary values to run on your cluster.