Whole-Genome Sequencing and Structural Variant Calling

Genomic DNA (1 μg) was randomly sheared to 350-bp fragments with Covaris cracker (Covaris) followed by sequencing library preparation using the Truseq Nano DNA HT Library Prep kit (Illumina). Sequencing libraries were sequenced paired-end 150 bp on the HiSeq X Ten sequencing platform (Illumina) with the HiSeq X Ten Reagent Kit v2.5 (Illumina) to a mean depth of coverage of about 50x. Reads were mapped to GRCh38 genome assembly using BWA-0.7.17 [53 (link)] with the default BWA-MEM parameters and 24 threads. SAM files were processed to sorted and indexed BAM files using SAMtools [52 (link)]. 2GS simulated reads were mapped to GRCh37 genome assembly using Minimap2 with the “-ax sr” parameter. For SV calling with novoBreak [26 (link)] (version 1.1.3rc), sorted and indexed BAM files were input with default run parameters. A dummy BAM file was simulated (GRCh38) to be used as a matched normal control. The confidence score for each breakend was obtained from the QUAL scores in the output VCF file. For SV calling with Delly [27 (link)] (version 0.7.8), duplicated reads in the BAM files were identified by Picard MarkDuplicates [54 ] before running Delly with the provided hg38 exclude file and its default parameters.