Author: Zechen Chong
Email: chongzechen@gmail.com or kchen3@mdanderson.org
Draft date: Aug. 15, 2017
novoBreak is a tool used in cancer genomic studies to discover SV (both somatic and germline) breakpoints. It can report accurate breakpoints of Deletions (DEL), Duplications (DUP), Inversions (INV) and Translocations (TRA) (you should consider some of them are mobile elements insertions or templated insertions). For novel insertions, we may only report the breakpoints but not the inserted sequence. Please forget about novel insertions at the moment. We will work on that later. novoBreak was designed for Illumina paired-end data.
Please cite novoBreak:
Chong, Zechen, et al. "novoBreak: local assembly for breakpoint detection in cancer genomes." Nature Methods (2016) PMID: 27892959.
novoBreak runs on a x86_64 Linux system with a ~40GB physical memory. It depends on SSAKE for local assembly, bwa-mem for contig mapping, and samtools (v1.3 and above) to extract reads. I have already put them in the release. If you cannot run the dependencies, please download the lastest versions of these tools.
git clone https://github.com/czc/nb_distribution.gitThen, please also add this directory to your PATH:
export PATH=$PWD/nb_distribution/:$PATHType
novoBreakYou should get the following output:
novoBreak - a tool for discovering somatic sv breakpoints
Auther: Zechen Chong <zchong@mdanderson.org>
Version: 1.1 (r20151007)
Usage:
novoBreak -i <tumorbam> -c <normalbam> -r <reference> -o <output.kmer> [options]
Options:
-h This help
-i <string> Tumor bam file
-c <string> Normal bam file
-r <string> Reference file in fasta format
-k <int> Kmer size, <=31 [31]
-o <string> Output kmer
-g <int> Output germline events [0]
-m <int> Minimum kmer count regarded as novo kmers [3].Otherwise, please install novoBreak first
git clone https://github.com/czc/novobreak_src.git
cd novobreak_src && makeReplace novoBreak in nb_distribution with the compiled novoBreak
Preprocessing: For virus integration analysis, please add the virus genome(s) to reference and realign the reads to the new reference. The integration should be a TRA event. Input files: a reference sequence file and two bam files (tumor and normal) of mapping results of the paired-end reads Output file: a filtered high confident VCF file (novoBreak.pass.flt.vcf). I left the intermedium files. You can delete them or inspect them as you need.
run novoBreak: The shell script for easy (hopefully) run of novoBreak is in the release directory. You can tune the parameters as you wish.
bash <A_PATH>/novoBreak/run_novoBreak.sh <novoBreak_exe_dir> <ref> <tumor_bam> <normal_bam> <n_CPUs:INT> [outputdir:-PWD]To increase sensitivity, novoBreak tries to infer as many SVs as possible based on the local assembly results. But many of the inferred SVs may be false positives due to misassembly or lack of enough evidence. So we provided a default filter to get a relatively stringent filtered callset based on real data experience. We empirically defined the minimum SV size as 100 bp and no upper limit. Users can change the filter and cutoffs based on the utility and the knowledge as needed. An empirical filter can be made based on the column 6 of novoBreak's output. A higher value of column 6 indicates a more reliable event. But sometimes it can be real event when breakpoints fall in the repetitive regions but with a low value. A detailed column information of the VCF files (*.sp.vcf) before applying filter is below:
Besides the standard 10 fields, the VCF file contains 29 additional fields can be used to build your own filters. These fields (column 11 to column 39) are: cluster_id, contig_id, contig_size, reads_used_for_assembly, average_coverage, tumor_bkpt1_depth, tumor_bkpt1_sp_reads, tumor_bkpt1_qual, tumor_bkpt1_high_qual_sp_reads, tumor_bkpt1_high_qual_qual, normal_bkpt1_depth, normal_bkpt1_sp_reads, normal_bkpt1_qual, normal_bkpt1_high_qual_sp_reads, normal_bkpt1_high_qual_qual, tumor_bkpt2_depth, tumor_bkpt2_sp_reads, tumor_bkpt2_qual, tumor_bkpt2_high_qual_sp_reads, tumor_bkpt2_high_qual_qual, normal_bkpt2_depth, normal_bkpt2_sp_reads, normal_bkpt2_qual, normal_bkpt2_high_qual_sp_reads, normal_bkpt2_high_qual_qual, tumor_bkpt1_discordant_reads, normal_bkpt1_discordant_reads, tumor_bkpt2_discordant_reads, normal_bkpt2_discordant_reads. Here, 'bkpt' represents 'breakpoint'; 'sp' represents 'split'; 'qual' stands for 'quality'.
Thanks for the help of the Bioinformatics team from Seven Bridges, we have optimized and released novoBreak as a public workflow (https://cgc.sbgenomics.com/public/apps#ZCHONG/novobreak-commit/novobreak-analysis/). Please feel free to use and we welcome any feedback from the users. Thanks!