Fig. 1: The workflow for benchmarking alignment- and assembly-bassed long-read SV callers. | Nature Communications

Fig. 1: The workflow for benchmarking alignment- and assembly-bassed long-read SV callers.

From: Tradeoffs in alignment and assembly-based methods for structural variant detection with long-read sequencing data

Fig. 1

The inputs are PacBio Hifi, Pacbio CLR and ONT datasets and their subsampled datasets at different sequencing coverages. The whole benchmarking workflow includes 12 read alignment-based SV calling methods, 4 assembly-based SV calling methods, 4 aligners, and 7 assemblers. It involves several different downstream analyses to evaluate all these SV callers. Assembly-based methods are highlighted in green, deep learning-based methods are highlighted in purple, the hybrid method is highlighted in cyan, and traditional alignment-based methods are not highlighted.

Back to article page