Proposed Analysis: Create tsnes with molecular subtype highlighted for Fig S4

jharenza commented 3 years ago

What are the scientific goals of the analysis?

I drafted some figures for Figure S4 to highlight some relevant subtypes within the tsne plot. This PR should update the tsne or figures code to generate these figures for the paper.

What methods do you plan to use to accomplish the scientific goals?

For this, I removed all "To be classified" subtypes
I think we should only look at EPN, MB, HGG, and LGG
For HGG, add both DMG and HGG samples
For LGG (panel not shown because I didn't update code for this), we can use all samples subtyped within the LGG module (LGG, GNG, GNT), but remove the prefixes LGG/GNG/GNT and only show subtypes so that we don't have an overload of so many groups. This will allow us to determine if we see clusters by pathway/subtype.
We could also make the NA in the legend be Other CNS tumor but still color code it as grey.

What input data are required for this analysis?

RNA-Seq stranded, pbta-histologies

How long do you expect is needed to complete the analysis? Will it be a multi-step analysis?

?

Who will complete the analysis (please add a GitHub handle here if relevant)?

someone at CCDL cc @jaclyn-taroni

What relevant scientific literature relates to this analysis?

NA

jharenza commented 2 years ago

@runjin326 can you work on this?

runjin326 commented 2 years ago

@jharenza , sure! will be able to get to this early next week.