KD-BIO
Large-Scale Data Science and Artificial Intelligence Infrastructure for Knowledge Discovery in the Green Life Sciences
KD-BIO is an integrated infrastructure for applying data science and artificial intelligence (DS&AI) to understand, model, and predict complex biological systems across organisational levels in the Green Life Sciences. It connects and complements Dutch infrastructures such as NPEC, UNLOCK, WI-KNAW, ELIXIR, the Jan Ingenhousz Institute, Holomicrobiome, and CropXR. KD-BIO enables domain scientists and DS&AI experts to jointly develop advanced, data-centric methods that uncover biological processes currently obscured by complexity and scale. By integrating high-dimensional phenotypic, environmental and management data with mechanistic models, it supports hypothesis generation, AI-assisted experimental design, and predictive cultivation and breeding. FAIR data management and semantic technologies ensure optimal reuse of datasets. KD-BIO serves academic, industrial, and societal users by providing accessible tools and expertise for data-driven research and decision-making, and by establishing non-invasive, scalable approaches to infer microbiome-mediated functions from plant phenotypes and to dissect the genetic architecture of complex traits such as photosynthetic efficiency.
KD-BIO accelerates scientific and technological innovation for scientific breakthroughs in the green life sciences through AI-driven, data-intensive research. By integrating diverse, rich datasets with mechanistic biological knowledge, it reveals how genetic, environmental, and management factors interact to shape complex traits under real-world conditions. This improves phenotype prediction and supports data-driven breeding, agronomic management, and microbiome-informed interventions. These advances directly contribute to major societal challenges: climate adaptation of agro/horti-production systems, sustainable, resource-efficient food production, nutritious food security in a biodiverse world. KD-BIO embeds Open Science and FAIR data, strengthening Dutch and European leadership and ensuring broad reuse of tools and knowledge.
Infrastructure and User Services
The KD-BIO infrastructure will offer the following:
- User-friendly Virtual Labs and Research Environment: These tools simplify research processes, enable easy searching and workflow composition, and foster collaboration among researchers.
- Standardized Data Science and Artificial Intelligence (DS&AI) Components and Workflows: This service provides generic and domain-specific data components and workflows, enhancing interoperability and reusability. Advanced AI-driven models offer accurate predictions and insights, aiding in better decision-making.
- FAIR Data/Workflow Processing: Training programs and data processing services help the scientific community manage and analyze large datasets responsibly, promoting open science and collaboration to accelerate discovery.
- Data Storage and Processing Capabilities: Robust services for storing and processing large data volumes assist researchers in expediting insights and knowledge discovery.
- Digital Testing and Experimentation Services: A digital environment for standardizing AI-driven model comparisons, utilizing resources such as HPC and scalable storage, streamlining model testing, and fostering rapid development of research.
- Training Sessions: Educational programs on DS&AI technologies.
- Life Science Consultancy and Technical Support: KD-BIO provides consultancy on GLS research problems, ethical considerations, and technical support for users.
Examples/description of the type of research
KD-BIO will empower Green Life Sciences (GLS) researchers with a comprehensive suite of advanced Data Science and Artificial Intelligence tools, enabling the integration and analysis of multidimensional datasets generated across GLS infrastructures. It will enable scientific breakthroughs through state-of-the-art data-centric methods and their integration with mechanistic approaches, and will support the design, reuse, and execution of data-centric and mechanistic workflows. This will help answer the fundamental scientific questions (e.g., What are the fundamental design principles of complex biological systems? What are the mechanisms and consequences of the responses of biological systems to their dynamic environments? How can we apply this knowledge to restore, improve, and design both natural and cultivated systems?) and drive the scientific breakthroughs required to predict ecosystem development in response to global change, design economically and environmentally sustainable circular biosystems, and guide plant production and breeding to create future high-yield crops adapted to changing and more dynamic and challenging climate conditions.
KD-BIO will promote interdisciplinary collaboration and enhance the precision and efficiency of research workflows, thereby accelerating scientific discoveries and driving innovations in the GLS, that can be adopted by industry. The infrastructure’s training and capacity-building activities will enable scientists to responsibly utilise cutting-edge AI technologies and adopt the FAIR data principles. Ultimately, this infrastructure will allow researchers to predict the properties and behaviour of complex dynamic systems, uncover hidden patterns and feedback loops, and deepen our understanding of natural systems. It will drive impactful advancements in the GLS, and inspire new scientific questions previously considered unanswerable.
How to access the infrastructure
Access to the infrastructure follows three paths:
- New Use Cases (30%) – User groups propose use cases to the Steering Committee through annual calls. The Steering Board selects proposals. In the first five years, selection is based on impact and relevance in developing KD-BIO components.
- Academic, Industry, and Societal Users (60%) – Onboarding follows a sustainability plan, initially expanding the community and later ensuring long-term sustainability.
- Academic Partners (10%) – KD-BIO partners access the infrastructure for research and validation for free. The Steering Board makes access decisions based on priorities.