Analyze with DataSHIELD

What is DataSHIELD?

Some research projects need to pool data from several studies to get sample sizes large enough to detect interactions. Ethical and legal constraints often prevent sharing individual-level data across studies.

DataSHIELD addresses this: it enables advanced statistical analysis of individual-level data from several sources without pooling the data together.

In collaboration with the DataSHIELD team at Newcastle University, OBiBa built the software infrastructure required to run DataSHIELD analyses securely on data stored in Opal. See the Opal R and DataSHIELD user guide for implementation details.

Features

With DataSHIELD in Opal, studies can:

• Create large virtual pooled datasets from multiple harmonized study datasets and run analyses such as linear regressions.
• Install and manage DataSHIELD R packages from the OBiBa CRAN repository.
• Set fine-grained user permissions on the tables and variables available for DataSHIELD analysis.
• Set up authentication certificates between DataSHIELD clients and Opal.

A real deployment is described in the BioSHaRE user story.