search

jbloomlab / SARS-CoV-2-RBD_DMS

Deep mutational scanning of the receptor-binding domain of SARS-CoV-2 Spike
BSD 3-Clause "New" or "Revised" License
43 stars 17 forks source link

Compute binding and expression scores, and fit global epistasis models #27

Closed tylernstarr closed 4 years ago

tylernstarr commented 4 years ago

Alright, the snakemake pipeline now proceeds all the way through global epistasis fits for the binding and expression phenotypes.

Getting the .Rmd files to render from snakemake/the command line (as opposed to "knitting" within RStudio) was a bit of a pain, as dealing with placement of files in subdirectories during the render apparently causes many people lots of headaches -- so as you'll see in the Snakefile, I settled on just doing everything in the base directory and then moving the .md and images to results/summary after rendering.

I have partially interpreted the binding global epistasis models, but I haven't annotated the md text in the expression global epistasis models as much yet -- but I wanted to just get this stage PR in since a lot was happening.

Also, I am committing my environment.yml file. I did rebuild my conda env from environment_unpinned.yml several times when trying out different R builds, though in the end I ended up reverting to the original environment_unpinned.yml -- but my environment.yml file appears to remain different from what was on master previously so I have kept that in the PR.

Also not sure what conflict arose with master commits which is preventing automatic merging of this PR, but I will also look through the change log here...