For a selected premature stop, there can be many candidate designs and several reasons why a promising molecule may not work in a cell. KritRNA is developing a platform to help make candidate choice more informed, explainable and experimentally actionable.
Making a candidate is one challenge.
Choosing the right one is another.
We are developing two connected engines to help make that choice more informed. One focuses on the candidate. The other considers the translation system it enters. Both are designed to learn from experiments.
Engine 01 is being developed to explore suppressor-tRNA candidates in the context of a selected nonsense mutation. It brings molecular suitability and the intended restoration goal into a common decision process.
Message availability, translation, quality control and protein output influence one another.
Engine 02 is a developing systems model of translation. It asks how a candidate may interact with message availability, ribosome behaviour, termination, quality control and the production of useful protein.
Explore candidate possibilities for a defined mutation and intended protein-restoration goal.
Assess readthrough alongside protein output, function, cellular response and effects at normal stop codons.
Develop experimentally challengeable prioritisation with a clear rationale and explicit uncertainty.
Use each research programme to inform the next while recognising that delivery and disease biology remain programme-specific.
The engines support a design–test–learn programme. Functional protein restoration, delivery and safety are experimental questions, not outputs that a model can certify.
The intended starting point is a defined gene, nonsense mutation and protein-restoration goal. The intended destination is a scientifically reasoned shortlist of candidates for testing.
The engine is being developed to bring molecular suitability and biological context into the same conversation. It asks whether a candidate is plausible as a tRNA, whether its intended activity makes sense in the target context, and what evidence would strengthen or challenge that choice.
Why a design is worth testing, which observations support that decision and which questions remain open.
Prioritise experimental effort and make assumptions visible before resources are committed.
Use clearly attributed experimental outcomes to challenge earlier choices and inform the next round.
These public biological questions explain why translation therapeutics needs an integrated approach. Safety and useful protein function run across all four.
A tRNA must be recognised and charged by an appropriate aminoacyl-tRNA synthetase. Recognition of a stop codon alone is not sufficient.
The suppressor tRNA competes with the machinery that normally ends translation. The local sequence and the state of the ribosome matter.
Nonsense-mediated decay can reduce the available transcript. Its effect varies with the gene, transcript architecture and cellular context.
A promising molecular design still needs a credible route to the relevant tissue and an effective intracellular exposure.
Our ambition is a platform that earns trust through prospective evidence, traceability and meaningful experimental choices. These are development standards, not a claim of completed validation.
Start with how translation works, not an isolated sequence score.
Preserve the biological and assay context behind every measurement.
Make the rationale for an experimental shortlist understandable and challengeable.
Test whether prioritisation helps on new experiments, not only on data already seen.
Consider normal-stop effects and cellular responses alongside protein restoration.
Keep a traceable connection between a scientific question, the experiment and the next decision.
We are building a design–test–learn programme in which readthrough, full-length protein, protein function, transcript availability and cellular effects remain distinct questions. Scientific, assay, disease-model and delivery collaborations help turn those questions into evidence.
Build the evidence with us ↗