The Patient-centered Information Commons: Standard Unification of Research Elements (PIC-SURE) platform integrates different layers of clinical and genomic data from diverse data sources, providing a multifaceted approach to biomedical research.
The PIC-SURE platform was built on i2b2 (Informatics for Integrating Biology & the Bedside, a data model created for EHR data), with an Apache 2.0 license (open source). PIC-SURE has been deployed in both FISMA Moderate ATO and HI-TRUST environments.
PIC-SURE example user interface.
Features
PIC-SURE UI and API
The PIC-SURE platform provides both an intuitive graphical user interface (UI) and an application programming interface (API) to meet different use cases and levels of experience with data manipulation. The PIC-SURE UI allows for an investigator to search for variables of interest and to conduct feasibility queries. In this way, cohorts are built in real-time and results can be retrieved for analysis.
PIC-SURE UI |
PIC-SURE API |
|
Advantages |
Build queries using point-and-click interface to explore variables and aggregate counts |
Build specific queries using code and extract data directly into workspace |
Access point |
PIC-SURE website |
Didactic Jupyter notebooks in R and python, R Markdown files |
Building queries |
Query Builder tool |
Python and R functions |
Extracting data |
Data can be downloaded or exported to an analysis workspace |
Run query in python or R to export data to workspace |
Data |
Clinical and genomic data indexed at the variable level Patient-level curation and ingestion of each phenotypic variable and genomic variant Variable, table and study metadata ingested and indexed for search Decoded variables from all studies made available to the user for cohort filtering and export |
Query Builder
The Query Builder in the PIC-SURE UI allows users to interactively build queries and returns the counts of participants that meet the criteria. This tool functions similarly to a search engine, providing the ability to search a term of interest and returning variables that contain the search term. After determining a variable of interest, users can refine their query to filter participants with certain values and build a cohort. Both single and multiple criteria queries are supported.
GitHub
PIC-SURE is developed under an Apache 2.0 license; all code is open source and accessible on GitHub: https://github.com/hms-dbmi/pic-sure-all-in-one.
For users interested in the PIC-SURE API, we have developed several Jupyter notebooks (R and Python) and R Markdown files which contain PIC-SURE use-cases. These examples provide a starting point for users to get started with their own analyses, and are publicly available on GitHub: https://github.com/hms-dbmi/Access-to-Data-using-PIC-SURE-API.