Updates to `run-manuscript-analyses.sh` and samples --> tumor

[sjspielman]

Closes #1679

This PR performs a couple updates as follows:

• I noted in analyses/README.md that the tumor-purity-exploration module is Yes in the manuscript.
• Tumor purity analyses and figures had not yet all made it into scripts/run-manuscript-analyses.sh, so I have added any missing items there. I re-ran with the RUN_LOCAL=1 flag and that's all locally working. On the way I addressed a couple bugs:
  • In analyses/telomerase-activity-prediction/RUN-telomerase-activity-prediction.sh, the script 06-TERTp-EXTENDScores.R hadn't been called (womp!), so it's now called such that it is always run when this script is run.
  • This is going conflict soon with a similar change that is made in #1673, so just want to cue in @jharenza that we both had to make this change. We'll see who makes it into master first!
  • There was also one wrong path that made it into generate-figures.sh (needed a supp), so I added that here.
• I realized while working towards #1676 that we still have some terminology "samples" instead of "tumor" (see this manuscript issue from revisions: https://github.com/AlexsLemonade/OpenPBTA-manuscript/issues/368), so I went ahead and updated the relevant code and associated panels to make sure those are good to go for compilation. This included the following scripts with the given PDF outputs (might want to turn off whitespace while reviewing scripts for some EOL cleanups) -
  • fig1-sample-distribution.R
  • 1B: figures/pdfs/fig1/panels/tumor_descriptor_proportion_panel.pdf
  • fig3_figS4-mutational-signatures-panels.R
  • 3E: figures/pdfs/fig3/panels/exposures_sina_IQR.pdf
  • S4A: figures/pdfs/supp/figs4/panels/exposures_per_sample_barplot.pdf
  • fig5_figS6-immune-deconv-panels.R
  • 5C: figures/pdfs/fig5/panels/quantiseq-cell_types-cancer_groups.pdf
  • 5E: figures/pdfs/fig5/panels/cd274_expression_mb_subtypes.pdf
  • S6C: figures/pdfs/supp/figs6/panels/quantiseq-cell_types-molecular_subtypes.pdf
• In the end, I re-ran generate-figures.sh and

[jashapiro]

It looks like in the current PR the updated figures are missing? (As in, there are a number of figures that have been deleted, which doesn't seem correct)

[sjspielman]

after confirming that the pdfs are really there despite some confusion about github notation.

It is indeed an interesting choice for github to show a red 100% indicator for files that are not 100% deleted, which the color red tends to imply!