Figs5 panels

[sjspielman]

This PR addresses issues (partially for SI) #1272 and #1281 and creates remaining S5 panels:

• figures/pdfs/supp/figs5/panels/supp_roc_tp53_polya.pdf: ROC tp53 from polya samples
• figures/pdfs/supp/figs5/panels/supp_terc_normextend.pdf: scatterplot of TERC FPKM across NormExtend scores ("telomerase score") figures/pdfs/supp/figs5/panels/supp_tert_normextend.pdf: scatterplot of TERT FPKM across NormExtend scores ("telomerase score")

Questions for reviewers:

• Should the telomerase scores label actually be "Normalized EXTEND scores" or should that just be in a legend?
• The original figures (created here) did NOT perform any normalizations or transformations (eg on FPKM), so neither did I. Still ok?
• I kept colors and trendline at default colors for scatterplots, instead of red points and black trend line. Any stylistic feedback?

Figures are documented in figures/scripts/README.md. Note that this PR and #1280 will deprecate the 04-telomerase-activities.R script (see Issue #1284).

[sjspielman]

@jaclyn-taroni updated this with log(FPKM +1). These plots look more reasonably scaled. Note that the correlations are performed with the transformed log(FPKM +1), so the p-values are influenced as well.