Establish Semantic Interoperability Across Research Using Ontologies
A case study conducted with one of the top 20 pharmaceutical companies worldwide.
The Goal: Increase research efficiency by use of uniform scientific nomenclatures across the entire research process and build a FAIR foundation towards future Data Governance and Knowledge Management activities
...faced by our customer:
Limited categorization of research data according to relevant entities and concepts
No capturing of contextualization and/or semantic information of life science data, metadata, concepts, terms and their relations
Limited number of controlled vocabularies and dictionaries with ambiguous definitions lacking internal alignment
A wide variety of ontology, taxonomy, and terminology management systems available on the market. Can they keep their promise in the complex life science domain?
Selection, validation, implementation and deployment of a state-of-the art ontology management system
Push standardized ontologies via APIs and REST interfaces to recipient lab and research informatics systems like ELN, chemical and biological registration, assay results, etc.
Exemplified data governance processes based on data foundation of the ontology management system
Increased awareness of customers‘ scientific community for the business values of FAIR data and „Data as an asset“ mindset
One system of truth for terminologies instead of many, reducing complexity and overhead of data management
New lab /research informatics systems can be adapted quickly to naming standards, reducing time to critical business insights
Semantic interoperability of research projects decreases ambiguity of scientific terminologies and its usage, increasing research efficiency and data quality
In the context of this case study, OSTHUS developed and supported an approach for enhancing data interoperability within the research and early discovery domain for one of the top 20 healthcare companies worldwide.
Biopharma and healthcare companies routinely encounter data interoperability challenges as this domain is amazingly complex, and controlled vocabularies, terms, and definitions needed to make the data interoperable are messy and / or ambiguous at best. This causes inefficiency in the data integration pipelines as the isolated activities to manage scientific vocabularies and taxonomies are bound to individual IT applications. One way to approach this problem is to develop a central repository for ontologymanagement.
Ontologies are structural frameworks for organizing information and are used as knowledge representation. If developed following industry standards, ontologies inherently provide a solid foundation to FAIR data and linked knowledge.
Life and physical sciences are more and more relying on such knowledge representations to cope with the complexity of the subject matter. Many large scientific ontologies have been developed and governed by expert communities, and are openly available to the scientific community. Ontology management platforms are information technology systems able to store and manage ontologies, and other structured information frameworks like taxonomies, terminologies, and dictionaries.
Typical features of ontology management systems include:
ingestion or building capabilities for ontology creation
the provision of standardized scientific terms and their synonyms
their corresponding definitions as well as
their relation to scientific concepts and the relationship between those concepts
search and editing capabilities across terms, concepts, and definitions
and the ability to provide these knowledge via modern interfaces like APIs or microservices
In many cases the introduction of ontology management systems is triggered by the desire to standardize namespaces, vocabularies, reference or metadata, to improve data management processes. All efforts aim at increasing the semantic interoperability and collaboration across the whole research organization.
The market sees a large variety of ontology management systems whose capabilities vary depending on their focus: managing terminologies, managing business and data assets, knowledge discovery or supporting data governance. Therefore, the introduction of an ontology management platform is often bundled with strategic efforts on an enterprise level to introduce data governance or knowledge management frameworks.
Change Management: Creating awareness for the potential and value of re-usable data
To introduce ontology management into the enterprise IT landscape, many investments have to be taken not only into the technology but also in people and culture.
Creating awareness for the potential and value of re-usable data and train scientists how to mine the information landscape are important pillars towards the digitalization of an organization.
The amount of training and change management required for moving to the mindset of using „Data as an asset“ is often underestimated in the data and IT strategy initiatives. Understanding this change needs to be prioritized when engaging with key stakeholders and subsequent scale-out activities.
Also, since every organization is different, this strategy needs to be tailored towards the one that is best-fit for the specific requirements, resulting in improved decision making, data quality and efficiency towards business and scientific insights.
The OSTHUS Consulting Approach
From vision to solution
Identified business use cases, assessed data sharing and standardization needs within the research organization of the customer
Identified technical and interface requirements for a respective solution within the customers’ IT organization. Summarized this as part of the requirements specification
Performed market analysis leading to the identification of several state-of-the-art ontology management system meeting said requirements
Orchestrated vendor-agnostic tender process for identifying the management system creating the biggest value for the customer's organization
Applied proof-of-concept study for validating and de-risking as said ontology management options. PoC pivoted around the demonstration of one or two use cases
Supported anchoring of selected technology provider in the deployment, integration and configuration of the ontology management platform
Incorporate relevant scientific ontologies
Assessed the maturity of internal terminologies with respect to knowledge capture and envisioned re-use in research IT landscape
Identified external domain-specific scientific ontologies overlapping with the research activities of the customer
Extended and applied ontology best practices (Pistoia Guidelines) to check the suitability of external ontologies for the scientific domain
Provided scientific advisory for upload/ publishing and data curation process within the central ontology management platform
Supported the life cycle management of internal and external ontologies within the organization
Solution: A state-of-the-art ontology management platform
Our Semantics and Life Science Informatics experts provided scientific advisory and knowledge sharing to stakeholders for creating awareness of ontologies and ontology management platforms to our customer's scientific community. This included the following:
Identification of a state-of-the-art ontology management platform meeting a large deal of formalized functional and technical requirements
Development and application of selection criteria for identifying external ontologies creating value for the company, following FAIR principles
Integration and annotation of more than 30 external ontologies relevant to the research organization’s scientific namespace
Establishment of Application Programming Interfaces (APIs) layer to surrounding / recipient systems to publish ontologies, namespaces, and concepts
Support of the life cycle of ontologies and data governance activities
Semantic interoperability through a centralized ontology management system
Results: From vision, to solution, to scientific insights
These are the core results, we were able to achieve for our customer:
Completed a solution alignment of the scientific and technical requirements within the customer's environment
Identified the best-fit and state-of-the-art ontology management platform in compliance with functional and technical requirements of the organization
Developed a detailed plan for implementation, deployment and integration of the said platform in the customer’s IT landscape
Combined internal nomenclatures with publicly available external life science ontologies to enhance contextualization, semantic interoperability, and knowledge discovery within the organization
Integrated the tool into the customer’s heterogeneous IT application landscape pushing concepts, terms, and metadata towards research relevant transactional and analytics systems
The ontology management platform created a single-source of truth for all scientific vocabularies, ultimately ensuring data consistency, FAIRness, improving search capabilities and trust in the data quality by all data consumers
Should you be facing similar challenges or you are simply looking to increase your research efficiency, our consulting experts are happy to find the ideal solution for your company.