BioModels: 15 Years of Powering Life Science with Code

The world's largest repository of curated computational models transforming biological research through standardization and reproducibility

Explore the Revolution

The Digital Revolution in Biology

Imagine trying to understand a city's traffic patterns by studying a single car. For decades, this was the challenge biologists faced: examining biological components in isolation while struggling to comprehend the emergent complexity of life itself. Then came a quiet revolution—the rise of computational modeling, where mathematical equations and computer code simulate biological processes from cellular metabolism to disease progression.

At the heart of this transformation lies BioModels, a groundbreaking repository that for 15 years has served as the digital library for life science models. Established in 2005 at the European Bioinformatics Institute, this platform has become the world's largest repository of curated computational models, emerging as the third most used data resource after PubMed and Google Scholar among scientists who use modeling in their research 1 7 .

15+
Years of Service
2,000+
Models Hosted
800+
Curated Models
60+
Organisms Covered

What is BioModels? The Digital Library for Life

At its core, BioModels is a specialized database for mathematical models describing biological processes. Think of it as a GitHub for biologists, but with rigorous curation and standardization that ensures every model actually works as intended. The repository hosts approximately 2,000 models from published literature, with about 800 meticulously curated models that have been verified to reproduce the results described in their original scientific publications 1 .

What makes BioModels revolutionary is its commitment to the FAIR principles—making models Findable, Accessible, Interoperable, and Reusable 1 . In a scientific landscape where published models were often lost or unusable due to format inconsistencies and insufficient documentation, BioModels provides a trusted platform where researchers can access reliable, semantically enriched models in standard formats that are easy to share, reproduce, and build upon.

Findable

Models are assigned unique, persistent identifiers and rich metadata for easy discovery.

Accessible

Models are freely available through standardized protocols without unnecessary barriers.

Interoperable

Models use formal, accessible, shared language and knowledge representation.

Reusable

Models are richly described with multiple accurate and relevant attributes.

The Curation Process: Ensuring Digital Reliability

The true value of BioModels lies in its rigorous curation process—a quality control system that transforms submitted models into verified scientific assets. This meticulous process follows the MIRIAM guidelines (Minimum Information Required in the Annotation of Models) and involves multiple stages of verification 1 6 .

Step Process Outcome
Submission Researchers submit models in various formats (SBML, CellML, MATLAB, etc.) Model receives unique submission ID
Format Conversion Models converted to standard formats like SBML Enhanced interoperability between software tools
Annotation Biological components tagged with controlled vocabulary Model elements linked to databases (UniProt, ChEBI, etc.)
Simulation Verification Independent reproduction of original publication results Certified reproducibility with different software tools
Publication Curated model assigned stable BIOMD identifier Public release with supporting documentation

The most crucial step involves verification that the model can reproduce published results—a process deliberately performed using different software than the original publication to eliminate tool-specific errors 6 . When simulations fail to match published outcomes, BioModels curators contact the original authors to resolve discrepancies, ensuring the digital reliability of every curated model.

BioModels Content Diversity

A Deep Dive into Model Curation: The Reproducibility Experiment

Methodology: The Verification Process

The cornerstone of BioModels' value is its experimental approach to verifying model reproducibility. This process transforms an unvetted mathematical model into a trusted scientific resource.

Model Acquisition and Standardization

The submitted model is first converted into standard SBML (Systems Biology Markup Language) format if necessary, ensuring it can be used across multiple simulation platforms 6 .

Independent Simulation Setup

Curators select a key figure or table from the original publication that demonstrates the model's dynamic behavior. They then recreate the simulation conditions described in the paper, deliberately using different simulation software than the original authors used 6 .

Execution and Data Collection

The model is run through the alternative simulation environment, generating output data for comparison with original publication results. Curators use tools such as COPASI, SBMLodeSolver, or Systems Biology Workbench rather than relying on the original author's preferred software 6 .

Comparative Analysis

Generated simulation results are quantitatively and qualitatively compared against the original publication's figures and data. Successful reproduction must capture the essential dynamic behaviors described in the paper.

Documentation and Annotation

The reproduced figure, along with curator comments on simulation parameters and software used, is permanently linked to the model for user reference 1 .

Results and Analysis: Ensuring Predictive Accuracy

This verification experiment produces crucial quality metrics for computational biology. When BioModels reports that a model is curated, it signifies that:

  • The model's mathematical formulation accurately encodes the biological processes described in the publication
  • The numerical implementation produces consistent results across multiple computational platforms
  • The dynamic behavior matches the reference publication within acceptable scientific tolerances

The reproduction success rate validates both the model itself and the broader ecosystem of systems biology tools. According to recent data, approximately 45% of kinetic models (548 out of 1,212 as of 2014) successfully passed this rigorous curation process to become fully MIRIAM compliant 6 .

Model Categories
Organism Distribution

The Scientist's Toolkit: Essential Resources for Computational Biology

BioModels provides researchers with a comprehensive suite of tools and standards that form the foundation of reproducible computational biology.

Research Reagent Solutions

Tool/Standard Function Application in Research
SBML (Systems Biology Markup Language) Standard format for encoding models Enables model exchange between 200+ software tools
MIRIAM Guidelines Annotation standards for models Ensures proper documentation and identification of model components
SBO (Systems Biology Ontology) Controlled vocabulary for processes and entities Adds semantic meaning to model elements
COMBINE Archive Container format for multiple related files Bundles models, data, and simulation descriptions
SED-ML (Simulation Experiment Description) Standard for describing simulation experiments Ensures reproducible numerical experiments

The web services provided by BioModels allow programmatic access to the repository, enabling researchers to search, retrieve, and analyze models directly from their computational workflows 8 . This API-based access facilitates the development of automated model analysis pipelines and complex multi-model studies that would be impractical through manual interaction alone.

SBML

Standard format enabling interoperability between 200+ software tools

MIRIAM

Annotation standards ensuring proper model documentation

SBO

Controlled vocabulary adding semantic meaning to model elements

The Future of Computational Biology

As we look ahead, computational modeling in life sciences is poised for even greater impact through integration with emerging technologies. The field is rapidly embracing artificial intelligence and machine learning, with life sciences executives increasingly investing in data science capabilities 5 . The convergence of mechanistic models from resources like BioModels with AI's pattern recognition capabilities promises to accelerate drug discovery and personalized medicine.

AI & Machine Learning

Integration of mechanistic models with AI's pattern recognition to accelerate discovery.

Digital Twins

Virtual patient replicas used for early drug candidate testing and personalized medicine 4 .

Quantum Computing

Potential application for molecular simulations and complex biological calculations .

Big Data Integration

Leveraging increasingly complex datasets for more accurate and comprehensive models.

BioModels' 15-year journey reflects a broader transformation in how we understand biology. By providing a trusted platform for sharing, verifying, and building upon computational models, it has helped shift biology from a purely observational science to a predictive one. As biological research continues to generate increasingly complex data, resources like BioModels will be essential for translating this information into genuine understanding and therapeutic breakthroughs.

In the words of one systems biologist, BioModels has become an indispensable tool, "emerging as the third most used data resource after PubMed and Google Scholar among the scientists who use modelling in their research" 1 . Its continued evolution will undoubtedly support future breakthroughs in our understanding of life's intricate mechanisms.

References