Final cleanup

[ErinWeisbart]

I'm merging in my changes as I consider them "final-draft". Before considering them truly final, they need

• [x] a formal code review by Beth
• [x] a read through by Meraj
• [x] to be run all the way through from start to finish to confirm everything is truly working as expected

I'm noting here a list of other cleanup that I know is necessary:

common_files

• [x] Barcodes.csv needs to be described

Hit_Calling

• [x] In README path to download mann_whitney.csvs is incomplete
• [x] In .ipynb need confirmation that gene lists are described correctly
• [x] Need plots added to outputs folder
• [x] Need mann_whitney.csvs uploaded to cellpainting-gallery

Figure 2

• [x] Document somewhere UMAPs change from run to run.
• [x] Save out the UMAP coordinates
• [x] The cluster labels aren't working on the UMAP
• [x] Using the latest CCLE version, the hit correlations dictionary crashes on frozenset({'TP53'}) for Fig2C

Figure 3

• [x] Using the latest CCLE version, the hit correlations dictionary crashes on frozenset({'MALT1'}) for Fig2C
• [x] HPLM UMAP isn't working
• [x] Save out the UMAP coordinates

Figure 4

• [x] Document how Fig 4A was made
• [x] Heatmaps don't obey font sizes passed to them so labels are funny sizes

Figure 5

• [x] x ticks missing

[ErinWeisbart]

@MerajRamezani I've updated Fig4 plots to have black along the diagonal in this commit

Basically, I'm adding another heatmap on top that is masked to only the diagonal and the diagonal values are set to be black.

ax = sns.heatmap(corr, mask = np.logical_not(np.diag(np.diag(corr))),linecolor='white', cmap = "mako", vmin=1, vmax=2,cbar=False)

I also cleaned up the axis labels a bit - there were ticks on one side but not the other and the x labels weren't center justified on their columns. Can you propagate these changes to other figures that have heatmaps? Thanks!

[MerajRamezani]

@mlozada21 make sure to add the description for Barcode.csv. I am currently working on the fig 4a.