Verix: a toolkit for benchmarking and harmonization of complex structural variants

Complex structural variants (CSVs) represent a critical yet under-characterized class of large-scale genomic rearrangements. CSV discovery requires not only the accurate detection of individual breakpoints but also their assembly into a single, structurally resolved CSV event. However, CSV method development is currently limited by a diagnostic gap: existing SV evaluation frameworks are designed for single-interval or single-junction SVs and cannot quantify multi-breakpoint CSV reconstruction fidelity or systematically distinguish between different types of reconstruction errors, such as over-segmentation or collapse of independent breakpoints. Here we present Verix, a comprehensive toolkit designed for the fine-grained evaluation and integration of CSVs. Verixintroduces granular diagnostic metrics that enable researchers to thoroughly characterize CSV predictions and stratify performance based on breakpoint sensitivity, precision, and structural integrity. In addition to pairwise callset-to-truthset comparisons, Verixalso provides an intra- and inter-callset harmonization module for generating consensus calls for a single tool's output or across disparate discovery pipelines. As such, Verixprovides the infrastructure needed to rigorously evaluate and reconcile CSV calls, accelerating the standardization of CSV discovery in both research and clinical settings.

Verixis available under the MIT license at https://github.com/PopicLab/verix and https://doi.org/10.5281/zenodo.22998063.

Supplementary data are available at Bioinformatics online.