Fig. 2
From: Genomic variant benchmark: if you cannot measure it, you cannot improve it
A Number of citations for benchmark studies using Entrez submodule of the BioPython package [31]. B The percentage of reference genome that is included in different benchmark studies is shown for different benchmark datasets. Besides, the sample sizes are also depicted as point size, which vary for different datasets. Another difference across them is the variant types that are included. Some benchmark datasets characterize SNVs and indels, while others cover SVs as well. C The number of variants (SNVs, SVs, and indels) for each benchmark dataset is represented on the y-axis, while the publication PMID is displayed on the x-axis