Use Cases
What protein engineers use Fermvyne for.
Thermostability redesign for industrial fermentation temperatures. Non-natural pathway co-design with cofactor balance constraints. Directed evolution library pre-filtering before synthesis spend. Three problems. Concrete workflows for each.
Redesign for industrial temperatures
A specialty chemical manufacturer in Research Triangle Park needed a dehalogenase variant that could operate reliably at 65°C — the fermentation temperature their existing process requires. Their current enzyme was characterized at 37°C with Tm of 52°C. Directed evolution campaigns had stalled after 3 months.
Workflow with Fermvyne: submit the FASTA of the parent enzyme plus target Tm of 65°C and E. coli BL21 as expression host. The platform generates 12 candidate sequences, each scored for predicted Tm, expression yield category, and solubility. The top 3 candidates are ranked by a combined fitness score.
Order gene synthesis for the top 3 candidates (typically ~$300 from Twist or IDT). Express, run DSF for Tm validation. Full turnaround from request to synthesis order: 2 days. Previous approach: 8-12 weeks per round, 3+ rounds typically needed.
Non-natural route for bio-based 1,3-PDO
A university synthetic biology lab needed 4 enzymes for a non-natural 1,3-propanediol production route — a pathway that doesn't exist in any organism's native metabolism. The key constraint: cofactor balance. NADPH-dependent enzymes in steps 2 and 4 needed to be matched by NADPH-regenerating capacity in the overall route.
The pathway co-design endpoint accepts a route specification with substrate/product pairs for each step and cofactor constraints. Fermvyne co-designs all 4 enzymes simultaneously, checking stoichiometric NADPH/NADH feasibility across the route as a generation constraint. Output includes matched expression vectors and predicted compatibility scores for co-expression in a single E. coli strain.
Without pathway co-design, each enzyme would be optimized independently with no guarantee of cofactor compatibility. The co-design approach reduces the experimental combinatorial space from hundreds of potential combinations to a small ranked set.
Pre-filter before ordering synthesis
A CRO protein engineering team was running standard error-prone PCR campaigns: generate ~200 variants, express the library, screen for activity. Their hit rate — active variants showing improved activity — was running at 12%. Eight weeks per round. Three rounds to reach their activity target.
They used Fermvyne to pre-score their 200 variants before ordering synthesis. The platform ranks variants by predicted expression, solubility, and predicted activity (using the parent enzyme's substrate as context). They ordered synthesis only for the top 18 predicted actives from each library.
Fermvyne is not replacing directed evolution in this workflow — it's filtering the library before synthesis spend. The result: 72% wet-lab hit rate on the ordered set, compared to 12% before. Total campaign time: 2 weeks instead of 8, because a smaller, higher-quality set moves faster through expression and screening.
Your enzyme problem has a specific shape. So does our solution.
Describe your target reaction and we'll show you what Fermvyne can do with it.