Zenodo CSVs: Figure S3C,D,E and S5

[sjspielman]

Towards #1692 ~⚠️ Stacked on #1708~ Edit: All caught up!

Another PR of Zenodo CSVs, all in figures/scripts/.

Figure S3C-E: figures/scripts/figS3-panels-cn-chromothripsis.R

Figure S3

• One big CN heatmap export with gz compression 💪
• Two exports from chromothripsis boxplots. These plots were made from the same data frame, and I considered exporting just one file. But, then I decided that the "spirit" of this activity suggests that perhaps we should be explicit.. So, I filtered down the relevant data for each export.
• While I was in this script, I cleaned up some outdated comments and such about figure panel order
• Also I left you some code treats 🍬 !

Figure S5 all panels: figures/scripts/figS5-all-panels.R

Figure S5

• An ROC curve, straightforward.
• Another situation where we have the same data heading into two plots, so again I filtered down to relevant bits for each export.

[sjspielman]

Took a stab at different approach for the bin column names in https://github.com/AlexsLemonade/OpenPBTA-analysis/pull/1709/commits/d16a7f1e344ff36394109b8fad4b725361e22f7d. I reformatted names from # --> chrX_bin# where X is the chromosome it is part of, not the X chromosome! We could also use actual bin coordinates, as in chrX_start_stop_bin#... I'd sort of rather not, but could if needed.

[sjspielman]

Code does not run through CI, so merging before checks.