search

AlexsLemonade / OpenPBTA-analysis

The analysis repository for the Open Pediatric Brain Tumor Atlas Project
Other
99 stars 66 forks source link

Survival panel for figure4f #1325

Closed sjspielman closed 2 years ago

sjspielman commented 2 years ago

This PR partly addresses #1323 by adding a Figure4F panel forest plot showing results from the additive survival model with three terms: tp53 score, telomerase score, and hgg grouping (binary variable).

Points that are significant are filled black, and otherwise filled white. I tried to indicate that the HGG group is referring to HGG (as opposed to reference level of non-HGG), but not sure my labeling conveys this well. No matter what, I think this will have to be a textual explanation in the caption.

sjspielman commented 2 years ago

I can make something like this but it will take longer to make, likely by end of the week.

If preferred, we can probably use the survivalAnalysis plotting function you found to make this plot and not run into any dependency issues.

jharenza commented 2 years ago

I don't have a strong preference except that we have the info in the plot since that's the standard view. One other idea is printing a tidy table as a panel and assembling in illustrator, but that may have been your thought also.

sjspielman commented 2 years ago

I think a separate table will be tricky to precisely align to the plot, so I'm going to try out a pure ggplot2 strategy today!

sjspielman commented 2 years ago

I think I may have found a good strategy much faster than I worried it would take me! This is two panels combined with cowplot. The only thing I can't do is scoot in the annotation columns more to the left (but could be done in Illustrator I guess?). There's a bit of "hackiness" here adding spaces to make everything line up like a table, but I don't think those hacks are painfully apparent in the plot. forest_survival_tp53_telomerase_hgg_panel.pdf

jharenza commented 2 years ago

Looks great!!

sjspielman commented 2 years ago

@jharenza I was pleasantly surprised! Ready for approval?

sjspielman commented 2 years ago

I made a small code update to determine p-value labels with ifelse() instead of hardcoding, after separate conversations with @jaclyn-taroni.

sjspielman commented 2 years ago

Code changes do not require CI; merging before checks.