Open Collections

Abstract

Specimens and data sharing through biobanking activities has become increasingly important in global biomedical research. These practices offer the potential to accelerate scientific discovery by enabling the reuse of biospecimens and associated data. However, the benefits of data sharing can only be realized when data are findable, accessible, interoperable, and reusable (FAIR). Biobanks remain fragmented—each operating with different standards, data formats, access policies, and query interfaces—making it difficult for researchers to locate and integrate specimen-derived data. This thesis explores the application of semantic web technologies (e.g., RDF, SPARQL, and ontologies) to improve the discovery and integration of biobank data, particularly omics-derived datasets. In Chapter 1, I provide an overview of the evolution and significance of biobanks in biomedical research, highlighting their transition from physical repositories to data-centric infrastructures, and discuss the persistent challenges researchers encounter in locating, accessing, and integrating fragmented and non-standardized specimen-derived data. In Chapter 2, I examine the current challenges in biobank data discovery and integration, including the lack of standardized contextual metadata, inconsistent identifiers, and insufficient support for automated data reuse. I propose semantic web approaches as a viable solution to these problems by enabling machine-readable, interoperable metadata and flexible, federated querying. In Chapter 3, I present the development of the Next-Generation Biobanking Ontology (NGBO), an open-source application ontology designed to represent omics contextual metadata. It provides persistent identifiers and semantically rich metadata to support both human understanding and computational reasoning. Chapter 4 describes a proof-of-concept implementation of a semantic web application that enables federated querying of biobank metadata across multiple SPARQL endpoints. Chapter 5 offers a structured review of how ontologies have been applied in biomedical text mining, with the aim of clarifying the strengths, limitations, and open challenges of ontology-based approaches in this field. The review was undertaken to identify opportunities for leveraging text-mined knowledge in biobank data ecosystems and to inform recommendations for future applications. Chapter 6 presents the discussion, conclusion, and future directions, reflecting on the key contributions of the thesis and outlining recommendations for advancing semantic web technologies in biobank informatics.