gwas-template-services

Application to provide template related services for the deposition interface

Requires Python 3.6+

Installation - using docker

For information on docker see Docker documentation

1. Clone repository:

   git clone https://github.com/EBISPOT/gwas-template-services.git
   cd gwas-template-services

2. Build docker image:

   docker build -t template_services .

   Buildind an image named template_services

3. Run docker image:

   mkdir docker_logs
   docker run --detach \
          -p 8649:8000 \
          -v $(pwd)/docker_logs/:/application/logs \
          template_services

   • Detaching the running container for the shell
   • Mapping the exposed 8000 port to 8649 on the host
   • Mounting the docker_logs folder so the containerised app can write logs on the host
   • starting the template_services image

   When running the image, the application automatically started and is accessible on port 8649

4. Test if application is running:

   curl -X GET "http://localhost:8649/v1/template-schema" -H "accept: application/json"

   The expected out should be a JSON document listing the available schema versions:

   {
     "schema_versions": {
       "1.0": {
         "href": "http://localhost:8649//v1/template-schema/1.0"
       }
     }
   }

Installation - using conda

1. Clone repository:

   git clone https://github.com/EBISPOT/gwas-template-services.git
   cd gwas-template-services

2. Create conda environment:

   conda env create -f environment.yml

3. Activate environment:

   conda activate template_serv

4. Install template service packates:

   pip install .

5. Start the web application:

   gunicorn -b localhost:8080 app:app \
       --log-level=debug \
       --access-logfile=logs/access.log \
       --error-logfile=logs/error.log

Endpoints

/v1/template-schema (GET)

Endpoint to expose all available schema versions. Returns a JSON.

/v1/template-schema/{version} (GET)

Endpoint to expose schema definition for a defined schema version. Returns a JSON.

/v1/templates (POST)

Endpoint to generate a template spreadsheet in excel. Retruns a blob.

Parameters:

• curator - Describing if a user is member of the curator group or not (type: boolean, optional, default: false)
• summaryStats - if submitting full metadata (false) of summary statistics (true). (type: boolean, optional, default: false)
• prefillData - if data sent to the endpoint to prefill certain cells.

Structure of the pre-fill data:

{
  "sheetName": [
    {
      "column1": "c1_value1",
      "column2": "c2_value1",
      "column3": "c3_value1"
    },
    {
      "column1": "c1_value2",
      "column2": "c2_value2",
      "column3": "c3_value2"
    }
  ]
}

examples:

Generate template for curators to deposity full metadata set:

curl -X POST "http://localhost:9000/v1/templates" \
    -d  '{"curator" : true}' \
    -H "Content-Type: application/json" > template.xlsx

Generate template for a non-curator user, where studies don't have mapped traits, and associations are not shown:

curl -X POST "http://localhost:9000/v1/templates" \
    -d  '{"curator" : false}' \
    -H "Content-Type: application/json" > template.xlsx

Generate template for depositing sumamry stats for already published studies.

curl -X POST "http://localhost:9000/v1/templates" \
    -d  '{"summaryStats" : true}' \
    -H "Content-Type: application/json" > template.xlsx

Generate template for depositing sumamry stats for already published studies with pre-filled study metadata for an easier identification.

curl -X POST "http://localhost:9000/v1/templates" \
    -d  '{"summaryStats" : true, "prefillData" : {"study":[{"study_accession":"GCST002728","trait":"Yang-deficiency constitution","sample_description":"30 cases, 30 balanced constitution controls"}]}}' \
    -H "Content-Type: application/json" > template.xlsx

If pre-filled template is generated, the pre-filled values are password protected and users can only updated cells in columns were input is required.

Using the stand-alone scripts

Documentation comes later.