Somatic structural variants (SVs), which are variants that typically impact more than 50 nucleotides, play a significant role in cancer development and evolution, but are notoriously more difficult to detect than small variants from short-read next-generation sequencing (NGS) data. This is due to a combination of challenges attributed to the purity of tumour samples, tumour heterogeneity, limitations of short-read information from NGS, and sequence alignment ambiguities. In spite of active development of SV detection tools over the past few years, each method has inherent advantages and limitations. We aim to evaluate variables impacting our ability to accurately detect somatic SVs and further facilitate informative decision-making on important impactful factors. Using simulation studies, we evaluated single and combinatoric effects of SV caller, SV types and sizes, variant allele frequency (tumour purity), sequencing depth of coverage, and variant breakpoint resolution. Using a generalized additive model allowed predictions of sensitivity and precision to be made for any combination of predictors. The prediction model was implemented in a web-based application, called Shiny-SoSV, which is freely available at https://hcpcg.shinyapps.io/shiny-sosv. Shiny-SoSV provides an interactive and visual platform for users to easily compare the individual and combined impact of different parameters. It predicts the performance of a proposed study design on somatic SV detection in silico, prior to the commencement of experiments. For details of previous and upcoming seminars, please visit https://www.maths.usyd.edu.au/u/SemConf/StatisticalBioinformatics.html