Build Process
Refreshing the OPS Platform
Producing a new version of the Open PHACTS platform requires both updating the software components and populating the software services with updated versions of datasets and linksets.
Core OPS software components
- Data Querying Service (i.e., Virtuoso + SPARQL)
- Identity Mapping Service
- Linked Data API Service
Steps to build (refresh) Data & Links
blue boxes are goal states
green boxes are actions
lines represent dependencies
Data Querying Service (Software)
Provides a SPARQL Endpoint for querying pharmacological & life sciences data.
Build & Deploy Process:
- Select Triple-Store product & version (e.g., Virtuoso)
- Obtain (download)
- Build and/or install
- Configure
- Integrate
- Run
- Expose/Shield
- Dockerize
- Publish
- Document
- Maintain
Data Querying Service (Data)
Obtaining, preparing, and loading RDF data into the Data Service triple-store (i.e., Virtuoso)
-
Dataset identification, inventory, selection, location
-
Pre-Download data file preparation
-
Download of source data files
-
Unpack archives
-
Transform non-RDF data into loadable RDF data files.
-
Assembly of all required files into a staging area.
-
Add supplementary files.
-
Load the RDF data into the triple-store (Virtuoso)
-
Execute post-loading SPARQL Update queries.
-
SPARQL Data Service final steps
-
Dump loaded data for future “quick loading”.
Identity Mapping Service (Software)
Obtaining, building, configuring, running IMS - Identity Mapping Service
IMS consists of 4 software components:
- QueryExpander
- IdentityMappingService
- Validator
- BridgeDb.
Plus 2 database components:
- MySQL
- Derby
Build & Deloy Process:
- Obtain (download/clone from GitHub)
- Build and/or install
- Configure
- Integrate
- Run
- Expose/Shield
- Dockerize
- Publish
- Document
- Maintain
Identity Mapping Service (Data)
Create, obtain, organize, integrate, and load the linkset data to populate the IMS.
Build process:
-
Linkset identification, inventory, selection, location
-
Pre-Download linkset preparation
-
Download Linkset files
-
Unpack linkset archives
-
Transform into IMS-ingestible format.
-
Generate linksets from software.
-
Assembly of all required Linkset files into a staging area.
-
Add supplementary files
-
Load IMS
-
Post-loading IMS final steps.
-
Dump loaded data for future “quick loading”.
Linked Data API Service (software)
Obtaining, building, configuring, and running the Linked Data API REST web service.
Optional/Supplementary components
- Explorer
- KNIME
- ops.js
- ops search
- Java library to access Linked Data API.