Studies Rest API Endpoint to Get All Studies

[sprintell]

As developer I want an API endpoint such that when I send a GET request to the URL, hateoas compliant, paginated list of all studies in the database is returned in JSON Format.

This will be a replacement to the old endpoint https://www.ebi.ac.uk/gwas/rest/api/studies

{
  "initialSampleSize": "...",  
  "replicationSampleSize": "2,698 European ancestry Bipolar disorder cases, 1,649 …",
  "gxe": false,
  "gxg": false,
  "snpCount": 1922309,
  "qualifier": null,
  "imputed": true,
  "pooled": false,
  "studyDesignComment": null,
  "accessionId": "GCST000854",
  "fullPvalueSet": false,
  "userRequested": false,
  "pubmedId": "20971583",
  "platforms": [{}],
  "diseaseTrait": "Suicide risk",
  "genotypingTechnologies": [{}],
  "efo_uri": "EFO_0004229",
  "efoTrait": "Dupuytren Contracture",
  "_links": {
    "self": {
    "href": "{base_url}/v1/studies/GCST000841"
    },
    "ancestries": {
    "href": "{base_url}/v1/studies/GCST000841/ancestries"
    },
    "search": {
    "href": "{base_url}/v1/studies?pubmed_id"
    }
  }
}

Must Have Additional Payload Attributes - These are displayed on the catalog UI:

• discovery_ancestry (country of recruitment): "4390 European (U.K.)"
• Traits: ["mental or behavioural disorder", "attempted suicide"]
• Platform [SNPs passing QC]: "Affymetrix [1922309] (imputed)"
• Replication ancestry (country of recruitment): "1649 NR (NR), 2698 European (Germany)"
• Full Summary Statistics: "NA"
• Terms/Licence: "https://www.ebi.ac.uk/about/terms-of-use/"
• backgroundEfoTraits: ["",""]

Management Rules

• Rule 1: Endpoint - /v1/studies
• Rule 2: HTTP Request Method - GET
• Rule 3: Endpoint should accept search page and sort parameters as in /v1/studies?page=0&size=20&sort=…
• Rule 4: Endpoint should be accessible to all
• Rule 5: HATEOAS, sorting, pagination, and some search and filter specification
• Rule 6: Endpoints should be cached based on request parameters to amplify query speed

API Endpoint Search Parameters This search parameter will automatically eliminate the need for 10 redundant endpoints as it was in the old design, listed below. Therefore the /v1/studies endpoint should accept search filter parameters as in /v1/studies?{search_parameter}=

Search Parameters:

• pubmed_id
• disease_trait
• fullPvalueSet
• userRequested
• efo-uri
• efoTrait
• accessionId
• fullPvalueSet

Few additional details here

[sajo-ebi]

@sprintell need some clarity on the new fields in API

[sprintell]

@sajo-ebi and @sprintell should investigate in addition of how to enhance this rest end point to handle child trait data ...