collect Opal 2.14 Released

We are pleased to announce that Opal 2.14 is now available. Opal is OBiBa’s core database application for biobanks.

Thanks to the funding of the EUCAN Connect program, this release greatly improves the speed of the import and export of datasets, as well as their assignment to R. Thanks to reduced memory needs, much larger datasets can be handled (tested with a dataset of 200M data points and a 8G opal server).

This release also supports external identity providers based on the OpenID Connect protocol for a better integration with the hosting institution's authentication system. The implementation is based on the excellent Nimbus OAuth 2.0 SDK . This has been tested with Keycloak , but should work with any identity provider system exposing an OpenID Connect configuration discovery entry point.

Additional features implemented:

  • refresh of a project (reload data and meta-data from the database),
  • a project can have default settings (preferred export folder and identifiers mapping),
  • views can now be backed up and restored from the web interface (a feature that is still available using the Opal Python commands). The views restore action is also to be used when importing Opal Views downloaded from a Mica server (3.6),
  • support of SAS Transport, compressed SPSS, and compressed CSV data formats,
  • support of SPSS ranges and missing values specifications are handled at import and export,
  • the R option datashield.seed is systematically applied at DataSHIELD server-side R session creation. It is specific to an Opal instance and immutable and will be used by DataSHIELD server side functions for consistent random numbers generation,
  • DataSHIELD platform base classes (configuration and R parser) have been extracted from Opal into an external library, datashield4j , so that other data repositories can implement the DataSHIELD platform.

See Opal documentation for installation and operation instructions.