Wilms-06 update CNV methods: infercnv (2/3)

[sjspielman]

Towards #810

This is the secodn of three PRs which update the CNV-related steps of the cell-type-wilms-tumor-06 module. This PR both updates the 06_copykat.R script and 06_infercnv_exploration.Rmd notebook.

• I updated the 06_infercnv.R script as follows:
  • I made sure we actually use the threshold for identifying normal cells to use in the reference via the new argument score_threshold
  • I added a flag which can be used in CI testing. When run in CI, this script will attempt to use a reference set of normal cells, but we can't expect that the testing data actually has normal cells. So, turning on this flag will ensure no reference is used when the test data is used.
• I fixed a small bug in 06_infercnv_exploration.Rmd

---

The third PR will be filed once #872 is merged, since it will update code in the explore-cnv-methods.sh script which that PR established to use the new testing flag in 06_infercnv.R added here.

In addition, the third PR may need to somehow update the rendered html notebooks for the exploratory copykat and infercnv code which are stored, for the 5 explored samples in notebook/{sample_id}/. I cannot fully render these notebooks (though I confirmed they run with test data!) due to compute limitations, so I see these as the best two ways to handle them:

• We can actually potentially remove these notebooks from the repo and send them to results instead, since they are not part of the actual annotation pipeline. In this case, I would take those steps in the third PR.
• We can probably ask Maud to regenerate these notebooks once we merge all this code into main. In this case, I would file an issue for followup.

[sjspielman]

Note that this comment was removed because it is not accurate.

[jaclyn-taroni]

👍🏻