Revision: UMAP at tumor purity threshold

[sjspielman]

We'll want to re-generate the UMAP at the tumor purity threshold. For this, we'll add functionality to the transcriptomic-dimension-reduction module to:

• Generate UMAP for tumor purity filtered data
• Add a notebook to compare that UMAP to the original UMAP made from the full data

[jaclyn-taroni]

• Add a notebook to compare that UMAP to the original UMAP made from the full data

I’m not sure if you have to do that if you can skip right to #1665. Might save us some time.

[sjspielman]

Because this is how things work, I am naturally seeing this comment https://github.com/AlexsLemonade/OpenPBTA-analysis/issues/1667#issuecomment-1431960984 right after I finished writing a barebones notebook for this. I think I'll put it into the PR I file for this and during review we can decide to either:

• scrap the notebook and make it a script in figures/ instead
• have the notebook export a UMAP that can be copied over to figures/ for the #1665 figure