Dimensions Knowledge Graph

The Dimensions Knowledge Graph, powered by metaphactory, is a large-scale semantic knowledge graph solution designed for pharmaceutical and life sciences organisations. Built on more than 32 billion structured statements, it integrates global research knowledge with an organisation's internal data to simplify and accelerate the generation of actionable insights across the entire pharma value chain. It is intended for a broad range of users including lab scientists, data scientists, system biologists, infectious disease experts, immunologists, and bio-informaticians.

Data within the knowledge graph is drawn from the Dimensions interconnected scientific research database, public data sources such as STRING and UMLS, and rich domain ontologies that power precise and consistent data annotation. These ontologies capture entities extracted from scientific text as well as complex relationships — such as cause-and-effect relations between proteins and drugs — ensuring that all incoming data, whether internal or external, is semantically harmonised for discoverability, interoperability, and trustworthy decision-making.

Key Use Cases

• Fast-track target discovery and reduce research costs: Researchers can create and validate hypotheses against global research knowledge and literature, accelerating time-to-market for new drugs and therapies through company-wide access to interconnected internal and external data.
• Streamline processes from R&D to clinical trials to market access: Research and lab data can be made available to manufacturing plants, supporting knowledge transfer and the planning of clinical trials, manufacturing, distribution, and market access processes.
• Reuse existing data and knowledge: A unified semantic layer makes data FAIR, supports automation of data collection and integration, and enables reusability of data, ontologies, and vocabularies across the organisation.
• Speed up drug safety review and optimise risk control: Existing research findings can be leveraged as supporting metadata in regulatory review processes, accelerating approval timelines.
• Leverage knowledge graphs with LLMs: The knowledge graph can be combined with large language models and generative AI to scale business decisions with machine-generated insights and augment explicit knowledge with AI algorithms.
• Enable and accelerate complex discovery workflows: Data is transformed into actionable knowledge available company-wide, supporting research initiatives, business decisions, forecasts, and predictions.

What's Included

• Publication and research metadata: 143 million publications, 160 million patents, 21 million datasets, 7 million grants, and 1.9 million policy documents.
• People and organisations: 34 million researchers and 129 thousand organisations.
• Semantic annotations: 307 billion linked semantic annotations and 35 million research integrity trust markers.
• Pharma ontologies and vocabularies: 307 billion annotations covering 30 million concepts from 38 domain ontologies.

Key Opportunities with the Dimensions Knowledge Graph

• Augment explicit knowledge (symbolic AI) with AI algorithms.
• Support, augment, and scale business decisions with machine-generated insights leveraging global research information.
• Explicit knowledge delivered by symbolic AI provides trust and explainability for AI-generated outputs.
• Integration with generative AI and machine learning enables a shift from human-driven to human-in-the-loop decision models.

How It Works

• At the core is an explicitly defined and flexible semantic model that can be extended to incorporate internal data, built using metaphactory's semantic modelling interface.
• The model is based on open standards, supporting data reusability across use cases and domains within an organisation.
• End users access integrated knowledge through use-case-specific views for intuitive search, exploration, and visualisation, or via API for injection into existing BI or data visualisation interfaces.
• A graph pathfinder feature (beta) enables users to find paths and connections across the data.
• Built-in AI Assists deliver a layer of trust and explainability to AI-generated insights, helping scale business decisions.

Open Standards Foundation

• RDF Data Model: A flexible, open standard for data representation.
• Ontology Language: Formal definition of the domain model, extensible at any time.
• Vocabularies: Categorisation and classification of data as hierarchical structures.
• Rules and Constraints: Explicit cardinalities and constraints enabling automated reasoning.
• Query Language: A flexible query language to specify graph-shaped information needs.
• Linked Data Platform: Open standards to support FAIR data within the organisation.

The Dimensions Knowledge Graph is an all-in-one solution ready for integration with an organisation's existing data infrastructure and internal knowledge graphs. It has been demonstrated at scale — exceeding 33 billion statements on a GraphDB foundation — and supports trustworthy, explainable AI-based applications in conjunction with LLMs and generative AI across pharma use cases from target discovery through to clinical trials and market access.