Updated analysis: update Multi-Hit definition

[jharenza]

What analysis module should be updated and why?

oncoplot-landscape

What changes need to be made? Please provide enough detail for another participant to make the update.

I noticed many of the KIAA1549--BRAF positive samples are annotated as Multi-Hit. I am wondering if this is because of the presence of reciprocal fusions KIAA1549--BRAF and BRAF--KIAA1549 or if there are additional hits. I think that we need to separate out Multi-Hit from reciprocal fusions, and collapse these fusions, as a reciprocal fusion should only be counted once. If, however, there are multiple genes fused to one gene (which I am assuming is not the case here, but can be in other cases), then we might also consider a Multi-Hit-Fusion annotation so that we are delineating from Multi-Hit, which I usually take to mean distinct events (eg two different SNVs or an SNV+CNV or SNV+fusion).

What input data should be used? Which data were used in the version being updated?

same, v18

When do you expect the revised analysis will be completed?

~1 week?

Who will complete the updated analysis?

@cbethell or someone at CCDL?

[jharenza]

Actually, I do see Multi-Hit Fusion in the plot (missed that at first because I was looking for alpha order :P.

Looking at the TMB plot section with the spread of mutations, I am assuming many of those multi-hits are coming from CNV + fusion. At first, I was wondering if the CNVs were being added to the MAF and that is why they are appearing as multi-hits, but I see you are entering it in the cnTable argument of the read.maf() function. I have seen default behavior as below with maftools for CN, wherein CN is annotated as a smaller colored box within the variant box, and this may help us visualize the data better.

I was looking up the package definition and it states for cnTable:

Custom copynumber data if gistic results are not available. Input file or a data.frame should contain three columns in aforementioned order with gene name, Sample name and copy number status (either 'Amp' or 'Del'). Default NULL.

I wonder if we need to recode to Amp or Del in order to get this behavior. Maybe that will help here.

[cbethell]

Per #259 (comment) and the comments added at the top of the 01 and 02 ploidy notebooks of the focal-cn-file-preparation module,

The `pbta-histologies.tsv` file contains a `tumor_ploidy` column, 
which is tumor ploidy as inferred by ControlFreeC.
[Hence we] add a status column that defines gain and loss broadly 
to mirror what logic is in the ControlFreeC file.

Therefore, I am not sure whether or not we want to revisit this decision and recode these values in the CN file.

Perhaps we may want to recode for the purpose of plotting the oncoprint, where we perform the recoding in the 01-plot-oncoprint.R script as it does appear that this would allow for the behavior depicted in the example plot in the above comment. Any further thoughts here @jharenza?

[jharenza]

Perhaps we may want to recode for the purpose of plotting the oncoprint, where we perform the recoding in the 01-plot-oncoprint.R script as it does appear that this would allow for the behavior depicted in the example plot in the above comment. Any further thoughts here @jharenza?

I would be in favor of recoding just for the oncoplot within this module

[cbethell]

Perhaps we may want to recode for the purpose of plotting the oncoprint, where we perform the recoding in the 01-plot-oncoprint.R script as it does appear that this would allow for the behavior depicted in the example plot in the above comment. Any further thoughts here @jharenza?

I would be in favor of recoding just for the oncoplot within this module

Sounds good, I can get started on drafting a PR for this!

[jharenza]

thank you!