I created a ti_oscope repository but I'm having some issues trying to run it on a broad range of datasets.
Perhaps @lengning can chime in on this. Could you have a look at run.R. Does this seem like a reasonable way to run oscope, provided I have a counts matrix?
What about Oscope?
https://bioconductor.org/packages/release/bioc/html/Oscope.html