The benchmark results for cuteSV, Sniffles, and SVIM were still consistent with Exp1 and 2, while the performance of NanoVar was different from before. from Zenodo (doi: ), and evaluated the four approaches (Fig. Next, we downloaded the 5× data provided by C.Y.T.

1B (labeled as Exp2), and the performances of the four SV detection tools broadly agree with Exp1. To reproduce these results, we generated a new simulated dataset using VISOR (v1.0) with the previously provided commands ( ), and we evaluated the SV calling results using Truvari (version 2.1, which was used in our paper). 1A, referred to as Exp1) as a relative standard. Firstly, we presented the performance of each SV caller in our previous paper (Fig. Here we conducted four sets of parallel experiments using 5× simulated datasets (refer to Table 1) to reevaluate the performance of NanoVar (v1.3.8), and we included Sniffles (1.0.12), SVIM (v1.4.0), and cuteSV (v1.0.10) as controls. Furthermore, we welcome more experts and scholars in the scientific community to pay attention to our research and help us better optimize these valuable works. We hope that this commentary proves the validity of our previous publication, clarifies and eliminates the misunderstanding about the commands and results in our benchmarking. were due to them using another version of VISOR and Sniffles, which caused many changes in usage and results compared to the versions applied in our previous work. Furthermore, the errors proposed by C.Y.T. The robust benchmark results indicate that NanoVar has unstable performance on simulated data produced from different versions of VISOR, while other tools do not exhibit this phenomenon. To clarify these matters, we reproduced our previous benchmarking results and carried out a series of parallel experiments on both the newly generated simulated datasets and the ones provided by C.Y.T. wrote a correspondence claiming that the performance of NanoVar was underestimated in our benchmarking and listed some errors in our previous manuscripts. We published a paper in BMC Bioinformatics comprehensively evaluating the performance of structural variation (SV) calling with long-read SV detection methods based on simulated error-prone long-read data under various sequencing settings.

Clipboard Master 5.5.0.50921 download