Interrogation of Mammalian Protein Complex Structure, Function, and Membership Using Genome-Scale Fitness Screens
Code repository for reproducing analyses and figures in this report.
From the command line:
$ git clone https://github.com/robinmeyers/pan-meyers-et-al
This can be done a variety of ways.
$ cd pan-meyers-et-al
$ R
or if you use RStudio
$ cd pan-meyers-et-al
$ open Pan_Meyers_et_al.Rproj
Many of the scripts in this project take advantage of parallelization using the R package doMC. The number of threads used is customizable in the lib/globals.R file. Edit this file and set to an appopriate number. We recommend allowing 4GB of RAM per core (e.g. set 4 cores on a machine with 16GB of RAM).
From the R console, run the SetupProject.R script. This will install any missing packages and download data files from the figshare record, and do quite a bit of data munging. Follow the instructions when the program asks for figshare authentication.
> source("./R/SetupProject.R")
This will take a while to complete. Once it does, we recommend restarting your R session before continuing.
Note: downloading the data files from figshare for the first time requires a web browser to authenticate. If running on a remote machine without this capability, execute the following commands in an R console locally.
> library(rfigshare)
> figshare_article <- fs_details(6005297)
This will open a web browser session, log in through figshare, and generate a .httr-oauth file in your working directory. Copy this file to the project directory on the remote machine. Running the SetupProject.R script as above should no longer require the authentication in a web browser.
This R project uses a directory structure and functions from the R package, ProjectTemplate. run.project() runs each *.R file in the src/ directory.
From a fresh R session, run the following commands.
library(ProjectTemplate)
run.project()
This will take many hours to complete, depending on the number of CPU cores used. The bulk of this time is spent running the CORUM_Shuffle.R and huMAP_Shuffle.R scripts, which run the network permutation tests for all protein complexes in each dataset. Many of the analyses depend on output files from these scripts. However, any script in the src/ directory beginning with a lower number prefix (e.g. 00-, 01-, 02-, etc.) will not depend on these output files. Run these scripts individually with source() if you're short on time.