Assess BRAF fusion breakpoint distribution by ancestry

[rjcorb]

Purpose/implementation Section

What scientific question is your analysis addressing?

This PR assesses BRAF fusion breakpoint distribution by predicted ancestry, CNS region, and extent of tumor resection

What was your approach?

• Added breakpoint info columns to pbta ancestry histologies file
• Calculated frequency of common versus rare breakpoint and common breakpoint types by predicted ancestry
• Assessed association between breakpoint type and CNS region and extent of tumor resection

What GitHub issue does your pull request address?

#41

Directions for reviewers. Tell potential reviewers what kind of feedback you are

soliciting.

Which areas should receive a particularly close look?

Please review code in 02-fusion-breakpoints-by-ancestry.Rmd and ensure script runs successfully:

Rscript -e "rmarkdown::render('02-fusion-breakpoints-by-ancestry.Rmd')"

Is there anything that you want to discuss further?

I am currently pulling all breakpoint data for each patient rather than sample. Let me know if sample would be more appropriate

Is the analysis in a mature enough form that the resulting figure(s) and/or table(s) are

ready for review?

Yes

Results

What types of results are included (e.g., table, figure)?

What is your summary of the results?

• We observe that the majority of BRAF fusion breakpoints are one of the 4 common types.
• There is a significantly higher frequency of patients with the 15:09 breakpoint type in patients of asian ancestry (EAS and SAS).
• We also see highly significant associations between breakpoint type and CNS region, with patients with common breakpoints having a higher frequency of posterior fossa tumors. This is likely the reason for the significant association also observed between breakpoint type and extent of tumor resection, with common breakpoint tumors having a greater frequency of Gross/Near total resection.
• Among common breakpoints, patients with 16:09 breakpoint have a higher frequency of PF tumors and higher freq of Gross/Near total resections relative to patients with other common breakpoint types.

Reproducibility Checklist

• The dependencies required to run the code in this pull request have been added to the
  project Dockerfile.

Documentation Checklist

• This analysis module has a README and it is up to date.
• The analytical code is documented and contains comments.

[jharenza]

Can you add a run script with this module please

[jharenza]

Script 02 runs ok - will review again once 01 is working

[jharenza]

Note from slack: BS_SDZY6RZ3 this sample has two common breakpoints- inspecting arriba files, it looks like chr7:138867975 chr7:140787584 is the lower confidence one, so we should probably remove that one from analysis (I did this in #45 with a manual mutate making keep == no)

[jharenza]

@rjcorb i also realize I can remove a bunch of code from the next pr if we export the fusion results with breakpoint type, breakpoint here. That way it's not all duplicated. Do you want to export a data frame by patient id or add that onto the annotation file?

[jharenza]

Will approve this, noting the ENSEMBL has been down for a week with no timeline on getting it back up 😬

https://www.ensembl.info/2024/01/18/downtime-of-the-ensembl-search-and-asia-mirror/