One connected system — from microbial diversity to pilot-scale ingredients.
Demand for naturally derived ingredients used in nutraceuticals, cosmetics, flavours, fragrances and pharmaceuticals continues to grow. Yet many supply chains still rely on wild harvesting, long agricultural cycles and geographically concentrated sources. This can place pressure on biodiversity, create variable product quality and expose producers and consumers to supply, adulteration and safety risks.
Precision fermentation offers a more resilient route, but developing a new microbial production process can still take five to ten years. Current workflows are often fragmented and costly, and are developed one molecule at a time.
The project spans large-scale digital exploration of microbial diversity, AI-driven retrosynthesis, enzyme engineering, rapid prototyping across multiple microbial hosts, fermentation scale-up, and early economic, safety, regulatory and environmental assessment.
The ORIGIN Co-pilot links data and learning across the project, supporting traceable, evidence-based decisions throughout repeated Design–Build–Test–Learn cycles.
Reducing timelines from the current 5–10 years
Targeting more than 10 g/L and more than 95% purity
Alongside 30–50% lower greenhouse-gas emissions and at least 70% water savings per kg of ingredient
Applied from development through scale-up and de-risking
* Environmental figures are targets compared with relevant conventional production baselines. They will be evaluated through the project’s techno-economic and life-cycle assessment work and are not verified results.
Optimized Routing and Industrial Generation of Ingredients from Nature
ORIGIN
101289831
Horizon Europe
HORIZON-CL6-2025-01
HORIZON-CL6-2025-01-CIRCBIO-08
Innovation Action
48 months, from June 2026
€5,498,237
SilicoLife SA
7 partners · 6 countries
Seven connected objectives spanning discovery, design, experimental validation, scale-up, de-risking, exploitation and project integration.
Establish a high-potential portfolio of microbial enzymes for sustainable ingredient production through AI-assisted digital bioprospecting.
Create AI-optimised, host-aware biosynthetic pathway blueprints for 50 high-potential molecules.
Construct and test a diverse multi-host strain library and select the 10 best-performing molecule/host combinations.
Optimise the 10 lead strains and demonstrate three robust production processes at 30–100 L scale, targeting TRL 7.
Generate pre-regulatory safety, toxicokinetic, techno-economic and life-cycle evidence for the lead candidates.
Develop investor-ready product roadmaps, manage intellectual property, communicate results and engage citizens and stakeholders.
Coordinate the consortium and develop the ORIGIN Co-pilot as a shared intelligence, data-integration and learning layer.
ORIGIN explores a broad portfolio of natural molecules and biological diversity, then uses evidence-based filters to focus resources on the most promising candidates.
Click in the image to enlarge
Explore global microbial diversity to identify promising enzymes and biosynthetic gene clusters.
Use AI to design pathways and enzymes, then build and test them across multiple microbial hosts.
Optimise fermentation, demonstrate lead processes and evaluate their economic, safety, regulatory and environmental performance.
Translate results into impact, public engagement and routes to uptake, connected through the ORIGIN Co-pilot.
WP1 explores public and proprietary genomic resources to identify microbial enzymes and biosynthetic gene clusters with potential for natural-ingredient production. Sequence-, structure- and genomic-context methods help discover candidates beyond cultivated microorganisms, while provenance and Access and Benefit-Sharing requirements support responsible and traceable biodiscovery.
WP2 turns target molecules and enzyme candidates into ranked, host-aware biosynthetic pathway designs. Retrosynthesis, metabolic modelling and generative enzyme engineering are combined to identify feasible routes, anticipate bottlenecks and create build-ready blueprints for experimental testing.
WP3 translates digital designs into physical microbial prototypes. High-throughput automation and iterative Design–Build–Test–Learn cycles are applied across E. coli, Saccharomyces cerevisiae, Yarrowia lipolytica and Aspergillus niger to identify and improve the most promising molecule–pathway–host combinations.
WP4 develops robust fermentation and downstream-processing methods for the lead strains. Processes are optimised starting from small-scale bioreactors, and the three most promising candidates are advanced to pilot demonstration at 30–100 L, generating material and data for subsequent assessment.
WP5 evaluates whether the emerging processes are economically viable, safe, sustainable and suitable for future market entry. It combines early techno-economic screening, life-cycle assessment, pre-regulatory safety and toxicokinetic studies, and regulatory planning for Europe and the United States.
WP6 turns scientific and technical outputs into protected, visible and usable assets. It covers intellectual-property strategy, product roadmaps, policy briefs, communication and dissemination, stakeholder engagement and the NOVA-led Citizen Science initiative.
WP7 provides project coordination, governance, risk and quality management, data management and development of the ORIGIN Co-pilot. The Co-pilot is designed to connect structured knowledge across all work packages, support reproducibility and improve project-wide learning and decision-making.
Academic excellence, advanced research infrastructure, specialist SMEs and industrial biotechnology expertise — covering every stage from genomic biodiscovery and AI-enabled design to strain engineering, pilot-scale fermentation, safety and sustainability assessment, exploitation and public engagement.
Portugal
Coordinates ORIGIN and leads project management, the ORIGIN Co-pilot, exploitation, intellectual-property strategy and communication. SilicoLife also contributes AI-enabled pathway and protein engineering, microbial prototyping and integrated TEA/LCA workflows.
Denmark
Leads fermentation optimisation and pilot-scale production. DTU contributes its Pre-Pilot Plant, downstream-processing expertise and techno-economic and life-cycle assessment capabilities to translate lead strains into scalable processes.
Spain
Leads AI-assisted digital bioprospecting and biosynthetic gene-cluster mining. CSIC applies large-scale bioinformatics to discover and prioritise novel microbial enzymes while supporting sequence provenance and responsible biodiscovery.
United Kingdom
Contributes BaseData™, a large, globally sourced and Nagoya-compliant metagenomic resource, together with AI and computing capabilities for enzyme, protein-structure and biosynthetic gene-cluster discovery.
Switzerland
Leads AI-driven retrosynthesis and host-aware pathway and strain design. EPFL applies computational biology, metabolic modelling and AI tools to produce ranked, build-ready pathway blueprints for experimental validation.
The Netherlands
Leads AI-assisted strain engineering and high-throughput prototyping, with particular expertise in Aspergillus niger. dsm-firmenich connects automated strain construction and screening with industrial bioprocess development and scale-up.
Portugal
Leads economic, safety, regulatory and sustainability assessment and the Citizen Science initiative. NOVA also contributes generative AI for enzyme engineering, microbial prototyping, pre-regulatory testing, communication and societal engagement.
