[Paper Review] Semantic integration of disease-specific knowledge.

Anastasios Nentidis, Konstantinos Bougiatiotis+2 more|arXiv (Cornell University)|Dec 18, 2019

Biomedical Text Mining and Ontologies4 citations

TL;DR

This paper proposes a semantic graph pipeline using the iASiS Open Data Graph to automatically retrieve and integrate disease-specific biomedical knowledge from literature and ontologies. The approach generates unified, query-ready semantic graphs that enhance access to concepts, relations, and attributes for diseases like Lung Cancer, Dementia, and Duchenne Muscular Dystrophy, demonstrating its potential for scalable, dynamic knowledge integration.

ABSTRACT

Biomedical researchers working on a specific disease need up-to-date and unified access to knowledge relevant to the disease of their interest. Knowledge is continuously accumulated in scientific literature and other resources such as biomedical ontologies. Identifying the specific information needed is a challenging task and computational tools can be valuable. In this study, we propose a pipeline to automatically retrieve and integrate relevant knowledge based on a semantic graph representation, the iASiS Open Data Graph. Results: The disease-specific semantic graph can provide easy access to resources relevant to specific concepts and individual aspects of these concepts, in the form of concept relations and attributes. The proposed approach is applied to three different case studies: Two prevalent diseases, Lung Cancer and Dementia, for which a lot of knowledge is available, and one rare disease, Duchenne Muscular Dystrophy, for which knowledge is less abundant and difficult to locate. Results from exemplary queries are presented, investigating the potential of this approach in integrating and accessing knowledge as an automatically generated semantic graph.

Motivation & Objective

• To address the challenge of fragmented and hard-to-access disease-specific biomedical knowledge in rapidly growing literature and databases.
• To enable researchers to efficiently retrieve and unify knowledge relevant to a specific disease, including both common and rare conditions.
• To develop an automated, scalable method for generating disease-specific semantic graphs from heterogeneous data sources.
• To improve access to detailed information on disease concepts, including their relations and attributes, through a standardized semantic representation.
• To evaluate the approach across diverse disease contexts, including knowledge-rich and knowledge-scarce scenarios.

Proposed method

• The pipeline leverages the iASiS Open Data Graph as a foundational knowledge base for semantic integration.
• It automatically extracts and maps disease-relevant concepts, relations, and attributes from scientific literature and biomedical ontologies.
• A semantic graph representation is constructed by linking concepts based on their contextual and relational metadata.
• The system supports dynamic querying to retrieve knowledge in a structured, machine-processable format.
• The approach is evaluated through case studies on Lung Cancer, Dementia, and Duchenne Muscular Dystrophy to assess scalability and utility.
• The pipeline enables unified access to both high- and low-abundance knowledge by enriching sparse data with semantic relationships.

Experimental results

Research questions

• RQ1 Can a semantic graph pipeline effectively integrate heterogeneous biomedical knowledge into a unified, query-ready format for disease-specific research?
• RQ2 How well does the approach perform in retrieving and organizing knowledge for diseases with abundant versus limited literature?
• RQ3 To what extent can the semantic graph representation enhance access to detailed concept-level information, including relations and attributes?
• RQ4 Can the system support dynamic, scalable knowledge integration across diverse disease contexts, including rare diseases?
• RQ5 What is the potential of using the iASiS Open Data Graph as a foundation for automated, disease-specific knowledge curation?

Key findings

• The semantic graph pipeline successfully generated integrated, disease-specific knowledge graphs for Lung Cancer, Dementia, and Duchenne Muscular Dystrophy.
• The approach enabled structured access to concept relations and attributes, improving discoverability of relevant information.
• The system demonstrated utility in both knowledge-rich (Lung Cancer, Dementia) and knowledge-scarce (Duchenne Muscular Dystrophy) contexts.
• Exemplary queries confirmed the pipeline’s ability to retrieve and organize complex biomedical knowledge in a semantically coherent form.
• The integration of literature and ontology data into a unified graph representation enhanced the coherence and accessibility of disease-specific knowledge.
• The results support the feasibility of automated, scalable knowledge integration for biomedical research.