Zegami_scripts

SUMMARY OF WORKFLOW TO BUILD ZEGAMI COLLECTION

1. Clone the repository and cd into it.

   git clone https://github.com/jstitlow/Zegami_scripts

   cd Zegami_scripts

2. Add environmental variables for omero and PYTHONPATH to your bash profile

   nano ~/.bash_profile

   • export OMERO="/opt/OMERO.py-5.4.10-ice36-b105/bin/omero"

   • export HOST="omero1.bioch.ox.ac.uk"

   • export PYTHONPATH="/opt/OMERO.py-5.4.10-ice36-b105/lib/python/:$PYTHONPATH"

   • control + x; y

3. Begin OMERO session by typing the following command into the command line.

   make login

4. Generate zegami.csv file from OMERO.Figure names that include Zegami.

   python list-figures.py

5. Download figures as json files.

   python download-figures.py /path/to/outdir /path/to/zegami.csv

   • will have to delete first row because the code is looking integer values

6. Convert json files to jpg for the scoring app

   python figure-json2jpeg.py /path/to/outdir /path/to/zegami.csv

7. Convert .jpg files to .png for Zegami.

   python convert_jpg2png.py

8. Populate the zegami.csv file with datasets from Intermine, Davis lab, and the literature

   • Requires conda environment if running on the server

   • Create conda environment from the conda_py27_clone.txt file

     conda create --name py27 --file conda_py27_clone.txt

   • Activate the py27 conda environment

     'conda activate py27`

   python runFlymineQueries_zegami.py -outname -infile -query_dir -dataset_dir

   • can add additional data by adding .csv file to dataset_dir with FBgn's in 1st column for indexing

   • needs Py2 because of some dictionary nonsense

   • export PATH="/usr/people/bioc1301/miniconda3/bin:$PATH"

   • source activate py27

TO DO:

• add try to figure-json2jpeg.py script

  `File "figure-json2jpeg.py", line 47, in main

  if int(fig_json['page_count']) > 1:

  TypeError: string indices must be integers`

• merge the scripts

• fix dictionary issue in runFlymineQueries_zegami.py

KNOWN ISSUES:

### some of the figures fail to build with the figure-json2jpeg.py script

    * can see a print-out of the failed files in the command-line output
    
    * would be useful to log the output

### zegami.csv has to be cleaned up as follows

    * remove Maria's data and others that are mis-labelled
    * remove template data
    * remove hyphens from CPTI ID numbers
    * several typos in figure_name (CPTI numbers) that make it difficult to correlate
    * updated figures from same CPTI line are an issue

SUMMARY OF WORKFLOW TO USE SCORING APP

1. Generate image files (.jpgs) and ID table using the Zegami workflow above

2. Run questionnaire.py script

   python questionnaire.py /path/to/questions/file /path/to/answers/dir /path/to/image/dir