Can Split-read data for the Kremer et al. dataset be used directly as input?

[RacconC]

I am trying to reproduce the results of FRASER on the application of rare disease diagnosis and I downloaded the split read data from the link given in the article (https://doi.org/10.5281/zenodo.4271599). I found that due to structural issues, this data could not be used directly as input, but only merged. What can I do to make it available as direct input?