README.standalone

To run ginkgo on the command line you must:

1. Install ginkgo

  Simply follow the instructions in the README.md that comes with Ginkgo (make sure you update the
  value of main in both scripts/process.R and in scripts/reclust.R
  
2. Create a directory under the uploads directory in the ginkgo installation directory

  mkdir ~/src/ginkgo/uploads/new_run_123

  Note that the name of that directory is the $GINKGO_USER_ID. The main server uses a 20 digits
  random number. Feel free to use whatever you see fit.

3. Copy your compressed bed files (the output of bamTobed) into that directory

  cp *.bed.gz ~/src/ginkgo/uploads/new_run_123

4. Create a file with the list of cells. The file must be called "list":

  ls ~/src/ginkgo/uploads/new_run_123 | grep .bed.gz$ > ~/src/ginkgo/uploads/new_run_123/list

5. Create a configuration file with all the options for ginkgo. This is probably the trickiest point
   as little documentation exists for this file. However, by reverse-engineering the php code, I
   have built an example file with enough comments to be able to modify the most important aspects
   of the analysis.

  cp ~/src/ginkgo/config.example ~/src/ginkgo/uploads/new_run_123/
  
  # Edit the file with vi or your favourite editor:
  vi ~/src/ginkgo/uploads/new_run_123/config.example

6. Run ginkgo:

  cd ~/src/ginkgo/
  ./scripts/analyze.sh new_run_123