Add copy number alteration data

[chrisamiller]

This goes later for sure, as it gets complicated/messy

[chrisamiller]

I think this should be straightforward to do for per-gene copy number assignments. Just add a 'copynumber' db table with copy number and LOH calls per-gene for the TCGA data
- This file contains gain/loss calls: /storage1/fs1/timley/Active/aml_ppg/analysis/tcga/tcga.200/analysis200Samples/copyNumber/list.cnAltered.genes.dat
- This file contains LOH calls: /storage1/fs1/timley/Active/aml_ppg/analysis/tcga/tcga.200/analysis200Samples/copyNumber/list.updAltered.genes.dat
- you'll have to convert TCGA ids to UPNs using this table: /storage1/fs1/timley/Active/aml_ppg/analysis/tcga/tcga.200/translation.table.dat

Once that is done, we can do the same for the beatAML data. It does not have per-gene assignments already, but the copy number segments are here:
/storage1/fs1/timley/Active/aml_ppg/analysis/external_datasets/beat_aml/copy_number/cnvkit_segments_altered.tsv

A few bedtools commands and this annotation file should allow you to generate gain/loss calls per-gene (we do not have LOH calls for this data) /storage1/fs1/timley/Active/aml_ppg/analysis/annotation_data/b38/genes.blocked.protein_coding.ens95.bed. Talk to Sai or I about how to do those intersections

Don't worry about doing any copy number visualization, we'll incorporate that into the oncoprint for now, and then think about whether to do independent copy-number plotting later on.

[saimukund20]

Im working on this.