This is the R script used in the PCAWG-6 SV mechanisms project to group rearrangements into rearrangement clusters and footprints.

• gtools
• doParallel

This program takes a BEDPE format file as input and returns a BEPDE file with additional columns appended. Ideally the breakpoints of each SV should be at basepair resolution, i.e. the span or low and high end should be a single base. However, the script should work even when the breakpoints are fuzzy.

Note that the program will crash if the input BEDPE file is empty.

Usage: Rscript run_cluster_sv.R -bedpe <input.bedpe> Rscript run_cluster_sv.R -bedpe <input.bedpe> -chr <chrom.sizes> -cen_telo <centromere_telomere_coords> -out <output.prefix> -n <n.threads>

Named arguments: input.bedpe (-bedpe) - Input BEDPE file, without header line. chrom.sizes (-chr) - tab-delimited file with list of chromosome sizes. Default: ./hg38.chrom_sizes centromere_telomere_coords (-cen_telo) - tab-delimited file with list of chromosome sizes. Default: ./hg38_centromere_and_telomere_coords.txt output.prefix (-out) - Prefic for output files. Default: Input prefix n_threads (-n) - Optional INT for number of threads to use. Default: 1

Two files are output:

• <output_prefix>.sv_clusters_and_footprints - this file contains the first ten columns in the input BEDPE file with the following columns appended.
  • Cluster ID of the cluster of each rearrangement junction
  • Number of SVs in the current SV cluster
  • Footprint ID of the low end breakpoint
  • Footprint ID of the high end breakpoint
  • Start and end coordinates of the footprint of the low end breakpoint
  • Start and end coordinates of the footprint of the high end breakpoint
  • P-value of the last agglomeration step in this cluster
• <output_prefix>.sv_distance_pvals - distance P-values between each pair of SVs. Note that this is the raw P-value matrix, where columns correspond to anchor SVs (see the PCAWG-6 SV mechanisms manuscript for details). Inside the algorithm, the pairwise distances used for hierarchical clustering of SVs are derived from (M + t(M))/2, where m is the P-value matrix in this output file and t() is the transpose function.

The program scales quadratically with respect to the number of SVs in the, since the distance between each pair of SVs are computed. Thus, with

SVs with at least one breakpoint within the provided telomere and centromere boundaries are discarded. Therefore the output may have fewer SVs than the input. This feature was originally created to ensure that the output from this script are compatible with downstream analysis scripts in the PCAWG-6 SV mechanisms project...

The algorithm behind this script is described in the PCAWG-6 SV mechanisms manuscript.

Coming soon hopefully...

Copyright (c) 2014-2017 Genome Research Ltd.

Author: Yilong Li <yl3.at.sanger.ac.uk>, <liyilong623.at.gmail.com> Contributor: Helena Winata [email protected]

ClusterSV is free software: you can redistribute it and/or modify it under the terms of the GNU Affero General Public License as published by the Free Software Foundation; either version 3 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Affero General Public License for more details.

You should have received a copy of the GNU Affero General Public License along with this program. If not, see http://www.gnu.org/licenses/.