Fig. 5: Subsampling effect and across-datasets analysis of different SV callers. | Nature Communications

Fig. 5: Subsampling effect and across-datasets analysis of different SV callers.

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

Fig. 5

a–f Recall-precision-F1 curves show the subsampling effect on deletion and insertion SVs by read alignment-based tools on Hifi_L1. g–h Recall-precision-F1 curves show the subsampling effect on deletion and insertion SVs by assembly-based tools on Hifi_L1. The coverage depth varies from 5x, 10×, 20×, 30×, 40× to 50×. Solid lines with markers are for different coverage depths, and corresponding dashed lines are for genotyping (gt) accuracy. For deletion SVs, we zoom in on the top right part of the plot to demonstrate the curves more clearly. i Heatmap shows overall and genotyping (gt) F1 scores on 11 long reads datasets for 16 SV calling methods. Empty cells represent analysis that could not be performed (or finished within 14 days of runtime) for the tool in the corresponding row. Source data are provided as a Source Data file.

Back to article page