erikyao / repodb

Data Plugin

This repo is the Biothings plugin for repoDB v2.0-SNAPSHOT data. The requirements were discussed in smartAPI - Issue#85.

A sample document for Brolucizumab is shown below:

"repodb": {
    "drugbank": "DB14864",
    "indications": ⊖[
        {
            "NCT": "NA",
            "detailed_status": "NA",
            "name": "Exudative age-related macular degeneration",
            "phase": "NA",
            "status": "Approved",
            "umls": "C0271084"
        }
    ],
    "name": "Brolucizumab"
}

Side note: why we chose to manually upload the data files

Currently we use 2 data files:

1. full.csv, downloadble from https://unmtid-shinyapps.net/shiny/repodb/.
2. drugbank vocabulary.csv (note the space in the filename), which is from unzipping drugbank_all_drugbank_vocabulary.csv.zip (direct download link) from DrugBank Release Version 5.1.9 - Open Data.

The problem with downloading full.csv is that, every time the download button is clicked on the repoDB webpage, a session id is generated as part of the download link, like:

https://unmtid-shinyapps.net/shiny/repodb/session/9f5dcbef21859438b1c0cea784bde839/download/downloadFull?w=

We are not clear if the session id would expire or become invalid, so we chose not to use this kind of direct download links in the manifest.json.

For the drugbank vocabulary, the direct download link https://go.drugbank.com/releases/5-1-9/downloads/all-drugbank-vocabulary works well for browser users. It will be redirected to an Amazon S3 link like:

https://drugbank.s3.us-west-2.amazonaws.com/public_downloads/downloads/000/005/940/original/drugbank_all_drugbank_vocabulary.csv.zip?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=AKIAJTZC3DSCEEG75A6Q%2F20220802%2Fus-west-2%2Fs3%2Faws4_request&X-Amz-Date=20220802T214251Z&X-Amz-Expires=30&X-Amz-SignedHeaders=host&X-Amz-Signature=b9c857335459c67db6913bced5b41dcedcc2d6beefbd1f02e132c0d3896d87ae

However internally the hub will take the last compoent of the URL path, all-drugbank-vocabulary, as the filename, and because there is no suffix .zip in it, the hub will not know how to decompress it.