bws

BWS is a Django app to run web-services for BOADICEA.

Quick start

1. Installation:

   pip install -e git+https://[email protected]/CCGE-BOADICEA/bws.git#egg=bws
   

2. Update the bws.settings.py. In particular change the FORTRAN_HOME parameter to set the directory location for the cancer risk models. Depending on the file structure it may be necessary to also change the HOME location in BC_MODEL and OC_MODEL in this file. These and the PROBS_EXE and RISKS_EXE parameters define the location of the executables for the mutation probability and risk calculation:

   FORTRAN_HOME = "/usr/src/"
   ....
   
   #
   # BREAST CANCER MODEL
   BC_MODEL = {
       'NAME': 'BC',
       'HOME': os.path.join(FORTRAN_HOME, 'boadicea'),
       'PROBS_EXE': 'boadicea_probs.exe',
       'RISKS_EXE': 'boadicea_risks.exe',
   ....
   

3. If you do not have access to the vcf2prs module, then comment out the related imports in the bws.settings.py:

   #import vcf2prs
   #from vcf2prs import SnpFile, Vcf2PrsError
   

and change the get_alpha function to return as follows:

def get_alpha(ref_file):
''' Get PRS alpha from a reference file header. '''
 return
#    moduledir = os.path.dirname(os.path.abspath(vcf2prs.__file__))
#    ref_file = os.path.join(moduledir, "PRS_files", ref_file)
#    try:
#        snp_file = open(ref_file, 'r')
#        alpha = SnpFile.extractAlpha(snp_file.__next__())
#    except (IOError, UnicodeDecodeError, StopIteration):
#        raise Vcf2PrsError('Error: Unable to open the file "{0}".'.format(ref_file))
#    finally:
#        snp_file.close()
#    return alpha

4. If you need to start a Django project:

   django-admin startproject [project_name]
   

5. Add "rest_framework", "rest_framework.authtoken" and "bws" to your INSTALLED_APPS in settings.py:

   INSTALLED_APPS = (
       ...
       'rest_framework',
       'rest_framework.authtoken',
       'bws',
   )
   

6. Import the bws settings in settings.py:

   from bws.settings import *
   

7. Add web-service endpoints to the `urls.py:

   from bws import rest_api
   from rest_framework.authtoken.views import ObtainAuthToken
   from django.conf.urls import include, url
   ....
   
       url_rest_patterns = [
           url(r'^boadicea/', rest_api.BwsView.as_view(), name='bws'),    # breast cancer risk model
           url(r'^ovarian/', rest_api.OwsView.as_view(), name='ows'),     # ovarian cancer risk model
           url(r'^auth-token/', ObtainAuthToken.as_view()),
       ]
       urlpatterns.extend(url_rest_patterns)
   

8. Run tests:

   python manage.py test bws.tests.test_bws \
                         bws.tests.test_ows.OwsTests \
                         bws.tests.test_throttling \
                         bws.tests.tests_pedigree_validation
   

Note: To set-up the BWS web-services to use the polygenic risk score (PRS) and cancer risk factor components (i.e. used with BOADICEA v5 or higher) requires extra Django Permissions to be granted.

The run_webservice.py script can be used to submit requests to the BWS web-service. It takes a username and pedigree file and will prompt for the password and run the risk calculation via the web-service:

${PATH_TO_BWS}/bws/scripts/run_webservice.py --help
${PATH_TO_BWS}/bws/scripts/run_webservice.py --url ${URL} -u ${USER} \
                                             -p ${PATH_TO_BWS}/bws/tests/data/pedigree_data.txt

To run the conversion script

1. Open a terminal and go to the desired path (e.g. with cd).:

   cd ./to/desired/directory/
   

2. Clone (or download the project):

   git clone https://github.com/ulaval-rs/bws
   cd bws
   

3. Make a new Python environment (this assume that you have access to python from the terminal).:

   python -m venv venv
   

4. Install the project dependencies:

   ./venv/bin/pip install -r requirements.txt
   

5. Copy-paste the dev.env to .env. This will contain the environment variables needed to run the scripts.:

   cp dev.env .env
   

6. Fill the .env file with the correct information (REDCAP_TOKEN, REDCAP_URL and REDCAP_PEDIGREE_VARIABLE).

7. Run the retrieve.py file. This file retrieve the pedigree files (at specified REDCAP_PEDIGREE_VARIABLE) from REDCap and store them at ./data/pedigree/.:

./venv/bin/python ./bws/scripts/retrieve.py

8. If no error occurs, there should be .txt files at ./data/pedigree/.

9. Now to convert the .txt files to .csv files, run the following script:

   ./venv/bin/python ./bws/scripts/canrisk_2_csv.py
   

10. CSV files should be at ./data/new_pedigree/, and the merged CSV file at ./data/merged_pedigree.csv