add human variable gene exclusion list

[GWMcElfresh]

Hi everyone,

I'm analyzing some human data (along with the Lewinsohn group needing some mechanism to shield interpretation of mitochondrial gene variance) and these lists offer a starting point to the human gene exclusion list.

Do we want to include HLA and TCR genes as well?

[GWMcElfresh]

TCR genes added

[GWMcElfresh]

I've added Class I HLA genes, BCR variable/joining, KIRs, and the human TRB diversity gene. Class II HLAs can be upregulated via generally biologically relevant mechanisms, so I think it's justifiable to exclude them.

I think that covers it.

Some tests with cell typist started failing, but my changes should have nothing to do with any current tests, unless there's a fetch for gene sets matching the word "Exclusion". I'll look more closely at why these are failing.

[GWMcElfresh]

scGate and Celltypist both haven't updated in months, so there's no discernible (to me at least) for the tests to have passed in 86505f5, then fail in d001431 due to slightly different abundances in cell classifications, unless there's some leap day nonsense happening.

I updated the tests to adjust for this, but we may need to do a similar approach to the bioconductor versioning issues and give a tolerance if these tests don't pass (for whatever reason) on master.