Skip to main content
. 2020 Mar 9;11:159. doi: 10.3389/fgene.2020.00159

Figure 1.

Figure 1

Overview of the evaluation pipeline in this study. Minimap2 (Li, 2018) and NGMLR (Sedlazeck et al., 2018) were used to perform alignment. Minimap2 aligned against reference genome with parameters ‘–MD -x map-pb/map-ont -R “@RG\tID:default\tSM : SAM” -a’ and NGMLR with the default parameters. SVs were identified by five callers, including Sniffles (Sedlazeck et al., 2018), Picky (Gong et al., 2018), smartie-sv (Kronenberg et al., 2018), PBHoney (English et al., 2014), and NanoSV (Cretu Stancu et al., 2017). Sniffles detected all types of SVs with parameters ‘–genotype –skip_parameter_estimation –min_support 10’ and employed a novel SV scoring scheme to exclude false SVs based on the size, position, type, and coverage of the candidate SVs (Sedlazeck et al., 2018). NanoSV was set ‘-s samtools’ to detect SVs and used the clustering of split reads to identify SV breakpoint junctions based on long-read sequencing data (Cretu Stancu et al., 2017). PBHoney considered both intra-read discordance and soft-clipped tails of long read (>10, 000 bp) to identify SVs (English et al., 2014). PBHoney, smartie-sv, and Picky were used to identify SVs with default parameters.