OneHealth Nexus

INPIRATION

Emerging zoonotic diseases can arise through complex interactions between animals, humans, pathogens, and environmental factors. Researchers often work with fragmented field observations, scientific literature, historical cases, and public-health data from different sources.

This makes it difficult to connect evidence, identify overlooked relationships, recognise knowledge gaps, and determine which research questions should be prioritized.

This inspired us to build OneHealth Nexus, an AI-powered research intelligence platform designed to transform fragmented One Health observations into connected evidence and meaningful research priorities.

WHAT IT DOES

OneHealth Nexus provides a researcher-focused workflow for investigating emerging zoonotic threats.

Capture & Structure

Researchers can enter observations related to animals, humans, pathogens, environmental conditions, geography, and time.

AI converts this fragmented field information into structured research cases, making the information easier to analyze and connect with existing evidence.

Connect Evidence

Using Retrieval-Augmented Generation (RAG), the system retrieves relevant:

Scientific literature Historical cases Reliable public-health data

The retrieved evidence is connected to the corresponding observations, helping researchers understand the existing scientific context.

Discover Relationships

A One Health Knowledge Graph connects entities such as:

Species Pathogens Humans Environmental factors Geography Time Scientific evidence

This allows researchers to explore relationships that may be difficult to identify when information is stored separately.

Identify What's Missing

The system analyses the available evidence to highlight potential:

Knowledge gaps Conflicting evidence Understudied relationships

These insights can help researchers identify areas that may require further investigation.

Prioritize What to Study Next

OneHealth Nexus generates and ranks evidence-grounded research questions based on the connected evidence.

Each research question can include supporting sources and identified limitations, allowing researchers to validate and assess the AI-generated suggestions before using them.

HOW WE BUILT IT

OneHealth Nexus combines AI, Retrieval-Augmented Generation (RAG), Knowledge Graphs, and structured research workflows.

The core workflow is:

Observation → AI Structuring → Evidence Retrieval → Knowledge Graph → Knowledge Gap Detection → Research Question Prioritization

The platform uses a web-based researcher interface, an AI/LLM layer for information extraction and reasoning, RAG for evidence retrieval, and a knowledge graph for representing relationships across One Health domains.

The system is designed to assist researchers rather than replace scientific validation. Evidence, sources, and limitations remain available for researcher review.

CHALLENGES WE RAN INTO

One of the main challenges is handling fragmented and unstructured information from different sources.

Scientific evidence can also vary in quality, context, and reliability. Ensuring that AI-generated insights remain grounded in trustworthy sources is another important challenge.

Building meaningful relationships between species, pathogens, humans, environmental factors, geography, and time also requires careful data structuring.

Another challenge is identifying genuine knowledge gaps because a lack of available evidence does not necessarily mean that no research exists.

ACCOMPLISHMENTS WE ARE PROUD OF

We developed a concept that goes beyond conventional literature-search or surveillance tools by connecting multiple stages of the research process into a single workflow.

The core innovation of OneHealth Nexus is connecting:

Observations → Evidence → Relationships → Knowledge Gaps → Research Priorities

This approach helps researchers move from scattered information toward a structured understanding of emerging zoonotic threats and identify areas that may deserve further investigation.

WHAT WE LEARNED

Through developing OneHealth Nexus, we learned how different AI technologies can work together to support scientific research.

We explored how RAG can ground AI-generated responses in external evidence, how knowledge graphs can represent complex relationships, and how structured information can improve AI-assisted research analysis.

We also learned that AI-generated research insights should be treated as decision-support outputs requiring human validation, particularly in scientific and public-health applications.

WHAT'S NEXT FOR ONEHEALTH NEXUS

The next step is to develop a functional prototype covering the complete research workflow.

Future improvements could include:

Integration with larger scientific literature and public-health databases More advanced Knowledge Graph construction Improved evidence-ranking mechanisms Temporal and geographical analysis Better detection of conflicting and missing evidence More advanced research-question prioritization Researcher feedback and validation mechanisms

Ultimately, OneHealth Nexus aims to become an intelligent research companion for investigating emerging zoonotic threats, helping researchers move from fragmented observations to connected evidence and research priorities.

Innovation

Unlike conventional surveillance or literature-search tools, OneHealth Nexus brings the complete research workflow into one researcher-focused system:

Observation → Evidence → Relationship Discovery → Knowledge Gaps → Research Priorities

From fragmented observations to connected evidence and research priorities.