could I run the sciclone to find subclones from mouse WES data?
It is not a totally absurd idea, take a look
I don't see why that wouldn't work, as long as there are enough
somatic variants to track from the tumors. (Cluster confidence increases with number of variants and the depth of sequencing).