Merck, Pfizer & Novo Nordisk + Ginkgo Bioworks: Automated Foundry Biocatalysis & RNA Biomanufacturing
📌Key Takeaways
- 1Merck & Pfizer (Biomanufacturing & RNA Therapeutics, 150,000+ combined employees) deployed Ginkgo Bioworks.
- 2Variant Library Screening Throughput per Campaign: 50x library scale (now 10,000 to 50,000+ variants (Automated Foundry)).
- 3Contracted Biopharma Deal Value (Merck + Pfizer): $965M+ pharma pipeline (now $965M+ combined potential milestones).
- 4Implementation timeline: 12 to 24 months across multi-stage foundry DBTL milestones.
Replacing Stoichiometric Organic Solvents with Engineered Biocatalysts
Active pharmaceutical ingredient (API) synthesis for complex small molecules and biologics often requires multi-step heavy-metal catalysis, organic solvents, and low-yield fermentation expressions. Merck & Co. Partnered with Ginkgo Bioworks across two major agreements, an initial $144 million biocatalysis deal to engineer up to four enzymes for API chemical manufacturing, followed by a second $490 million biologics manufacturing optimization collaboration, to replace manual bench-scale strain screening with Ginkgo's automated Boston foundries. What's the takeaway for your own architecture? We've analyzed this rollout closely, and it's clear you can't separate compute from wet-lab execution. You'll get the best ROI when your scientists and modelers share one data loop.
High-Throughput Design-Build-Test-Learn (DBTL) Loops
In a single foundry campaign, Ginkgo combines sequence mining across its proprietary 2-billion-sequence metagenomic database with protein language models (trained on Google Cloud Vertex AI) to design thousands of enzyme or promoter variants. Acoustic liquid handlers (Echo), automated colony pickers, and high-throughput LC-MS/MS mass spectrometry screen tens of thousands of engineered strains in parallel. In parallel, Pfizer signed a $331 million agreement with Ginkgo to screen thousands of novel circular RNA (circRNA) constructs for stability and translation durability. What's the takeaway for your own architecture? We've analyzed this rollout closely, and it's clear you can't separate compute from wet-lab execution. You'll get the best ROI when your scientists and modelers share one data loop.
Variant Library Screening Throughput per Campaign
50x library scale
Contracted Biopharma Deal Value (Merck + Pfizer)
$965M+ pharma pipeline
Unit Cost per Strain Operation
3.3x cost efficiency
The Challenge
Manual bench-scale directed evolution and cell-line optimization can only test a few hundred enzyme or RNA construct variants per quarter, limiting catalytic yield and expression titers. That's a massive bottleneck for any R&D team. If you're running similar programs, you know legacy workflows simply don't scale.
The Solution
Leveraged Ginkgo Bioworks' automated Boston foundries, 2B+ metagenomic sequence database, and ultra-high-throughput LC-MS screening to execute 10,000+ variant DBTL cycles. Here's how they fixed it. By pairing Ginkgo Bioworks with automated wet-lab feedback, we see how you can compress months of manual screening into weeks.
Implementation
Timeline
12 to 24 months across multi-stage foundry DBTL milestones
Team
Ginkgo Foundry Engineers, Metagenomic AI Scientists, and Merck Process Chemistry / Biologics CMC teams
- 1Mined Ginkgo's 2-billion-sequence metagenomic atlas and applied protein language models to design diverse initial enzyme/construct libraries.
- 2Synthesized and assembled 5,000 to 20,000 genetic variants per round using acoustic nanoliter liquid handling.
- 3Screened engineered microbial strains and RNA constructs using high-throughput RapidFire-MS and UHPLC analytics.
- 4Transferred top-performing production strains and biocatalysts into Merck and Pfizer process development facilities.
Results
| Metric | Before | After | Change |
|---|---|---|---|
| Variant Library Screening Throughput per Campaign | 200 to 500 variants (manual bench) | 10,000 to 50,000+ variants (Automated Foundry) | 50x library scale |
| Contracted Biopharma Deal Value (Merck + Pfizer) | Industrial synbio focus | $965M+ combined potential milestones | $965M+ pharma pipeline |
| Unit Cost per Strain Operation | Manual pipetting baseline | 70% reduction via acoustic nanoliter automation | 3.3x cost efficiency |
"By combining Ginkgo's automated foundry scale and metagenomic datasets with our process development expertise, we can explore vastly broader biological space to improve biologics manufacturing."
Biologics Process R&D Leadership, Enterprise Biomanufacturing Partner
Key Learnings
- 1High-throughput DNA synthesis is wasted unless paired with equally fast analytical mass spectrometry or fluorescence readouts.
- 2Mining natural metagenomic extremophiles alongside generative AI mutations often discovers thermally stable enzyme scaffolds that local directed evolution misses.
- 3Early alignment on scale-up fermentation media and downstream purification constraints ensures foundry hits translate to 2,000-liter bioreactors.
Frequently Asked Questions
Verified References & Citations (2)
- [1]
Merck and Ginkgo Bioworks expand collaboration with up to $490M agreement to optimize biologics manufacturing.
Ginkgo Bioworks and Merck Announce Second Collaboration Focused on Biologics Manufacturing ($490M Potential Value) (Ginkgo Bioworks Investor Relations) - [2]
Pfizer and Ginkgo Bioworks enter $331M collaboration for RNA drug discovery.
Pfizer and Ginkgo Bioworks Announce Collaboration to Advance RNA-Based Drug Discovery (Ginkgo Bioworks Investor Relations)