FAIR Assessment for Improved Accessibility of Protected APIs

Abstract

The inability to reliably share, reproduce, reuse and translate scientific data into usable information has been a known issue in scientific communities for some time. The National Institutes of Health (NIH) launched an initiative in 2013 known as Big Data to Knowledge (BD2K) to address these issues in light of an ever growing collection of biomedical data. The BD2K program funded efforts to build new tools that would specifically address gaps translating large data sets into useable information. One of those efforts was the Patient Information Commons - Standardized Unification of Research Elements (PIC-SURE) API whose goals included creating a toolset for querying patient level data across multiple disparate differential datasets through a single interface. But managing access to the APIs that grant users the ability to request protected data remains a top challenge. The success of PIC-SURE API and other BD2K efforts has now evolved into the Findable, Accessible, Interoperable, and Reusable (FAIR) Guiding Principles that outline a future where data resources can become auto-discoverable by both humans and machines. With the broad range of technical areas that the FAIR Guiding Principles outline, there have been many projects dedicated to presenting solutions that focus on improving biomedical data utilities. This project work aims to identify obstacles for API accessibility and demonstrate through a reference implementation how accessibility can be improved for a protected API. The proof of concept system leverages modern web technologies, open source community driven standards, and widely accepted communication protocols to present a system that reduces access barriers but still maintains a high level of security needed for protected resources.