Long-term preservation of biomedical research data.

Area of Science:

• Biomedical Science
• Data Science
• Information Science

Background:

• Biomedical science is increasingly data-intensive due to advances in genomics, molecular imaging, and translational research.
• Effective long-term preservation of Big Data sets is essential for researchers to leverage these resources.
• Current research data management practices often lack strategies for sustained accessibility post-project completion.

Purpose of the Study:

• To outline actionable strategies for ensuring the long-term preservation and continued resource value of biomedical research data.
• To advocate for the early involvement of data stewards in the digital data lifecycle.
• To promote the adoption of standardized practices for enhanced data sharing and integration.

Main Methods:

• Discussion of an opinion article proposing the use of the Open Archival Information System (OAIS) model.
• Emphasis on the six functional entities of the OAIS model: Ingest, Access, Data Management, Archival Storage, Administration, and Preservation Planning.
• Recommendations for data collection strategies, use of common data elements, and engagement with repositories and curators.

Main Results:

• The OAIS model provides a framework for managing digital data throughout its lifecycle.
• Early involvement of data stewards and adherence to institutional policies are key to data sustainability.
• Standardization through common data elements and ontologies improves data interpretation and reuse.
• Scalable platforms and secure virtual workspaces are needed to support diverse data ingest and facilitate FAIR data principles.

Conclusions:

• A proactive approach to data preservation, integrating data stewards early, is vital for biomedical Big Data.
• Implementing robust data management strategies, including provenance tracking and standardization, enhances data quality and reproducibility.
• Adoption of FAIR data principles through appropriate infrastructure is essential for near- and long-term research needs.