problems for detecting SVs

[azure-twilight]

I see in your work the results in detecting SV is pretty good as precision, recall and f1 score > 90%

I try to rereach the result. So as described in your paper, I use GRCh37 chromosomes 20-22 as reference, use pbsim to simulate CLR data, use SURVIVOR to simulate SV, use truvari to evaluate the results. But the results for minmap2/lra + pbsv/svim/cutesv is unexpected. For example, pbmm2 only get < 0.54 f1 score, 0.68 precision.

I don't know what's wrong with my experiment. Could you tell me your detail commands and parameters for the workflow in detecting SVs？

Thanks.