Add an "exclusivity/co-occurence" view for mutations

[chrisamiller]

What cbioportal calls an "oncoprint" - rows are mutations, columns are samples, boxes filled in/color coded if they're mutated. Consider adding stats for correlation/mutual exclusivity. Looking at what packages Glioviz uses for this might be helpful

[chrisamiller]

For Oncoprints, see the ComplexHeatmap package: http://www.bioconductor.org/packages/devel/bioc/html/ComplexHeatmap.html and the documentation here: https://jokergoo.github.io/ComplexHeatmap-reference/book/oncoprint.html

I don't know that it is the only solution (feel free to poke around for other tools), but it's one that seems like it could plausibly work.

See the image here for a decent example of what we have in mind:

The idea is that there would be a menu on the left that lets you select multiple genes, and each time you do, it adds a row to the oncoprint. The inputs to the oncoprint consists of the copy number data for a gene and the mutation data for a gene, so this depends on https://github.com/leylabdotorg/amlviz/issues/23

[chrisamiller]

Watch out for an occasional gene that may have both gain and loss assigned to it (the gain hits the front half, the loss hits the back half. Don't know if these packages handle that or it needs special handling from us!

[chrisamiller]

Gene fusions should be listed here as mutations. It is an open question whether we should add them to the mutations table or pull that data from the clinical table - we should think through the implications of those approaches