Background & Strategy

The pipeline integrates three interconnected modules:

Virus Evolution Algorithm

The pipeline begins by forecasting the evolution of the studied virus (i.e., H1N1 for this use case). By analysing already observed and forecasted strains, we will identify mutational patterns in the viral genome to identify regions that vary across strains and host animals, as well as those that remain conserved. This provides critical insights into pathogen stability, variability, and immune evasion. To future-proof the vaccine and maximise cross-strain coverage, conserved regions are prioritised in the subsequent epitope mapping step.

In Silico Epitope Mapping

NOI’s proprietary AI pipeline identifies regions of the pathogen's proteins most likely to trigger a protective immune response. The AI platform then optimises candidate epitopes for immunogenicity, manufacturability, and broad coverage across strains.

To ensure that only viable candidates advance, MD simulations model the behaviour and physicochemical stability of candidate epitopes under production and manufacturing conditions. The most promising antigens then undergo preclinical validation, including in vitro and in vivo studies, to confirm safety and efficacy across multiple models, establishing a proof of concept that serves as a benchmark for scalability and a foundation for future regulatory pathways.

Molecular Dynamics (MD) Simulations

Emerging infectious diseases and future pandemics continue to represent major global public health threats. PREPAIR-VAC addresses these challenges by integrating artificial intelligence technologies into vaccine production.

PREPAIR-VAC aims to create a fully automated, AI-driven pipeline that streamlines vaccine design and optimises manufacturing, significantly reducing the time needed to bring vaccine candidates to preclinical validation.

Experimental data generated during manufacturing and preclinical validation will continuously feed back into AI models to improve predictive performance and preparedness capacities.

Pre-clinical proof of concept

(PREPAIR-VAC project)

Pandemic scenario

Identified pathogen of concern

Genomic database of human & host animal strains

Core AI technology

Virus Evolution Model

In silico epitope mapping model

Epitope optimization

Optimized antigen candidates

Protein production for experiments

Pre-clinical validation (in vitro, in vivo)

Pre-clinically validated vaccine

AI-designed vaccines acceptance

Deployment

Scale-up

Translation

Translation

Core AI technology

Optimized antigen candidates

Pre-clinically validated vaccine

Large-scale manufacturing

Clinical development (clinical trial)

Policy & communication strategy

Vaccine roll-out

Beyond the science, PREPAIR-VAC also addresses the ethical, legal, and social dimensions of AI-driven vaccines through stakeholder engagement, ensuring that new technologies are trustworthy, unbiased, and ready for real-world deployment.

Objectives

Advance AI-supported approaches for viral evolution prediction and conserved epitope identification

Develop & optimise AI-powered antigen design and epitope prediction tools

Establish recombinant vaccine manufacturing workflows

Evaluate immunogenicity and protection profiles of AI-designed vaccine candidates through preclinical models;

Address ethical, societal and regulatory aspects related to trustworthy AI-supported vaccine technologies;