This project contains the links to the datasets and the code that was used for our study : "A benchmarking of WGS-based structural variant callers"
Table of contents
Sarwal, Varuni, et al. "A comprehensive benchmarking of WGS-based structural variant callers" bioRxiv, doi: https://doi.org/10.1101/2020.04.16.045120
We have evaluated 12 structural variant tools: Biograph, BreakDancer, CLEVER, DELLY, GASV, GRIDSS, indelMINER, MiStrVar, Pindel, PopDel, RDXplorer, LUMPY . Details about the tools and instructions for running can be found in our paper.
We have prepared "wrappers" in order to run each of the respective tools as well as create standardized log files.
The raw vcf's produced by the tools can be found here: https://github.com/Mangul-Lab-USC/benchmarking-sv-callers-paper/tree/master/Data/raw_data/mouse/raw_vcf The custom vcf files, which are raw vcf's converted to the VCFv4.2 format can be found here: https://github.com/Mangul-Lab-USC/benchmarking-sv-callers-paper/tree/master/Data/raw_data/mouse The fastq and bam files used will be available soon
The scripts to compare the deletions inferred by the SV-caller versus the true deletions is available here: https://github.com/Mangul-Lab-USC/benchmarking-sv-callers-paper/blob/master/Scripts/customvcf_mouse.py
We have prepared Jupyter Notebooks that utilize the raw data described above to reproduce the results and figures presented in our manuscript.
- Figure1 Jupyter Notebook
- Figure2 Jupyter Notebook
- Figure3 Jupyter Notebook
- Figure4 Jupyter Notebook
- Figure5 Jupyter Notebook
- Supplementary Figures Jupyter Notebook
This repository is under MIT license. For more information, please read our LICENSE.md file.
Please do not hesitate to contact us (mangul@usc.edu) if you have any comments, suggestions, or clarification requests regarding the study or if you would like to contribute to this resource.