Want more cell-selective protein expression and longer stability for your mRNA?

We’ve got you covered.
We don’t guess.
We optimize, while respecting the biology.
Every nucleotide position matters for patients.

The Challenges

mrna design is a balancing act.
Old methods below have fighting objectives.
True multi-parameter optimization has been hard, until we made our selectivity agent.

0
0
1
0
0
1
Strengthening structural stability can reduce targeting precision - losing cell selectivity.
Focusing on strengthening structural stability alone can reduce cell-type translation selectivity.
0
0
2
0
0
2
Product Strengthening structural stability can reduce targeting precision - losing cell selectivity.
Strict Kozak sequence requirements can reduce overall translational efficiency.
0
0
3
0
0
3
Tuning for cell selectivity frequently trades off protein yield - hurting expression.
Tuning for cell selectivity can sometimes introduce repetitive sequences or homopolymers.

These tradeoffs are not bugs: They are the natural balance of Biology.

Our nCode-optimizerTM handles this balance gracefully, and with >5x better performance on more objectives than traditional reinforcement learning (RL) post-training techniques.

Solution

Every nucleotide is optimized across different objectives, all at once.

5′ UTR (BPE) CDS (3-mer codon) 3′ UTR (BPE) [MASK]
Use Case

Cell Selectivity:
Target cells - high translation.
Off-target cells - suppressed translation.

mCherry mRNA optimized for translation efficiency in T-cells. 
All 5 top lead optimized candidates achieve around >5× selectivity across both off-target cell types (B-cells and hepatocytes).

01 — Lead candidates

High in target. Low everywhere else.

The top five leads express 5–6× more protein in T-Cell target cells than in either off-target cell type.

Image
02 — Population-level improvement

Every design shifts into the safe zone.

Each dot is one sequence. Optimized designs (teal) sit above the 1:1 line — high in target, low off-target — while base line sequences (blue-gray) stay on the diagonal.

Image
03 — Training convergence

Selectivity that holds.

Across training cycles the target reward climbs and stays high, while both off-target rewards stay suppressed.

Target (T-Cell) off-target (Hepatocyte) off-target (Mutu)
Let's optimize together!

What mRNA medicine payload do you want to optimize? In which target cells?

We adapt quickly to your goals;
in as little as one week.

Selectivity Agent

Worried about
cell selectivity, expression, and stability
Of your mRNA?

We've got you covered.We don't guess. We Optimize, while respecting the biology!
Every nucleotide and its position matters