search

AlexsLemonade / OpenPBTA-analysis

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

Re-run mutational signatures module with V22 #1476

Closed sjspielman closed 2 years ago

sjspielman commented 2 years ago

This PR re-runs the mutational signatures module, specifically by issuing CNS_FIT_ONLY=1 bash run_mutational_signatures.sh. Note that I had to clean up a bit of syntax in the run script due to weird spacing, and I fleshed out some of the comments in the run script for clarity.

There were several bugs that cropped up when re-running the notebooks that had to do with V22 changes, including how Oligodendrogliomas are handled and there appears to now be an additional hypermutator (BS_F0GNWEJJ - this could use some more eyes to confirm it's really supposed to be there!). As a consequence, I had to update the figures in this PR as well since that was the only way to address the bugs and get the overall module to run.

Overall, there are no strong changes in results except for the new (?) hypermutator sample. Note that I also added set.seed() to the sina and jitter plot code for reproducibility moving forward as this was missing.

jharenza commented 2 years ago

Great, I will take a look at this now @sjspielman . Note that we did see BS_F0GNWEJJ before, but this was an NBL metastasis and we chose to focus our discussion only on HGGs, so did not include. I thought we added that to the plot code in some way, but will take a look!

sjspielman commented 2 years ago

Note that we did see BS_F0GNWEJJ before, but this was an NBL metastasis and we chose to focus our discussion only on HGGs, so did not include.

Ah, this is coming back to me! I guess we're seeing now that we need to explicitly filter it out in this module.

jharenza commented 2 years ago

Note that we did see BS_F0GNWEJJ before, but this was an NBL metastasis and we chose to focus our discussion only on HGGs, so did not include.

Ah, this is coming back to me! I guess we're seeing now that we need to explicitly filter it out in this module.

In thinking more about this, I think we ought to include. I am scanning for predisposition and germline variants, but we will need to assess TP53 status in this sample.

sjspielman commented 2 years ago

I just have one question, and that is whether you want to update exposures_presence_barplot.pdf here since this PR reruns all figures and the text in the facet strips and the x-axis is a bit smushed.

Thanks for catching this, will do!

sjspielman commented 2 years ago

@jharenza I realize that, since this PR did all the mutational signatures figures too as a bonus treat, I should also copy them into figures/pdfs/! I'm re-requesting a review here to make sure those copies look ok to you!