Journal: bioRxiv

Article Title: OctopusV and TentacleSV: a one-stop toolkit for multi-sample, cross-platform structural variant comparison and analysis

doi: 10.1101/2025.03.24.645012

Figure Lengend Snippet: OctopusV is organized into three primary modules: the Input layer, the Correct module, and the Merge module, along with supporting functionality modules. The Input layer processes SV data from multiple sequencing platforms (NGS, PacBio, ONT, and others) and handles VCF files from standard SV callers. Several common callers are presented as examples; however, OctopusV can work with any VCF-formatted SV calls regardless of the caller or sequencing technology. The Correct module processes VCF files through sequential steps: BND extraction, pattern recognition of breakpoint orientations (shown by N[…] and […]N patterns), and SV type classification into standard forms (duplication, inversion, intra/inter-chromosomal TRA). The Merge module implements advanced merging strategies based on event coordinates and properties, supporting various operations including length overlap assessment, Jaccard index calculation, and set operations (union, intersection, specific). This module handles different SV types (DEL/DUP/INV, TRA, INS) with specific coordinate matching criteria (δstart, δend thresholds). Additional modules provide benchmarking, format conversion, and visualization capabilities. The output includes merged results in a customized VCF format and comprehensive visualization options (UpSet plots, SV size distribution, SV type distribution, and chromosome distribution). The customized VCF format serves as an intermediate representation that facilitates integration between modules. Additionally, OctopusV generates interactive HTML outputs that allow users to visualize and explore SV data through a web browser interface.

Article Snippet: The input layer handles variant call format (VCF) files from multiple sequencing platforms, including next-generation sequencing (NGS), Pacific Biosciences (PacBio), and Oxford Nanopore Technologies (ONT), and commonly used SV callers, such as Manta, LUMPY, SvABA, DELLY [ ], PBSV [ ], SVIM [ ], Sniffles [ ], CuteSV [ ], SVDSS [ ], and DeBreak [ ], converting them into a standardized format for analysis.

Techniques: Sequencing, Extraction

Journal: bioRxiv

Article Title: OctopusV and TentacleSV: a one-stop toolkit for multi-sample, cross-platform structural variant comparison and analysis

doi: 10.1101/2025.03.24.645012

Figure Lengend Snippet: a Workflow for benchmarking SV merging, illustrating processes from raw sequencing data to performance evaluation against reference datasets. The workflow includes SV calling to generate input VCF files, applying different merging strategies, and evaluating merged variants using Truvari. b Diagram illustrating SV merging operations supported by OctopusV. Common operations include intersection, union, and minimum support (variants supported by ≥2 callers). Unique operations provided by OctopusV include maximum support (variants supported by ≤2 callers), caller-specific extraction (e.g., Manta-only), and custom set logic operations. For NGS datasets (NA12878 NGS, VISOR NGS), the custom set logic operation is defined as ((Manta AND SvABA) NOT (LUMPY OR DELLY)). For PacBio and ONT datasets (NA12878 PacBio, VISOR PacBio, VISOR ONT), it is defined as ((DeBreak AND SVDSS) NOT Others). c–g Heatmaps comparing absolute Precision, Recall, and F1 scores of OctopusV against other SV merging tools (Jasmine, SURVIVOR, SVmerge, CombiSV) across multiple datasets: c NA12878 NGS, d NA12878 PacBio, e VISOR NGS, f VISOR PacBio, and g VISOR ONT. Cells marked with horizontal lines indicate unsupported operations; diagonal lines represent analyses with no results; dotted patterns indicate configurations not analyzed. The label "4callers" indicates analyses limited to the caller combination (Sniffles, PBSV, CuteSV, SVIM) specifically supported by CombiSV.

Article Snippet: The input layer handles variant call format (VCF) files from multiple sequencing platforms, including next-generation sequencing (NGS), Pacific Biosciences (PacBio), and Oxford Nanopore Technologies (ONT), and commonly used SV callers, such as Manta, LUMPY, SvABA, DELLY [ ], PBSV [ ], SVIM [ ], Sniffles [ ], CuteSV [ ], SVDSS [ ], and DeBreak [ ], converting them into a standardized format for analysis.

Techniques: Sequencing, Extraction