Our Approach


Our unique approach in small molecule mRNA drugs

Our unique platform combines high-scale phenotypic screening with AI mRNA image analysis, enabling us to discover selective mRNA small molecule drugs and rapidly elucidate their mechanisms of action. Our approach has resulted in a broad and quickly growing pipeline of promising candidates across a range of therapeutic areas and indications, which has been validated through our strategic collaborations with big pharma including Eli Lilly and Takeda.

Phenotypic screening has long been seen to have strong advantages in cases where the biology around a disease is not fully understood and targets are not yet validated. In such situations, a phenotypic screen represents an unbiased strategy that does not make an assumption about a specific target but instead, identifies molecules that actively modify cellular phenotypes in a disease model. One can find several chemical structures with such effect, giving projects, “multiple shots on goal”.

The challenge of phenotypic screens is when it comes to understanding the mechanism of action of the active compounds. It generates millions of colorful images of cells and identifies compounds that have caused modifications, but it does not provide much evidence as to what could be the reason for the observed changes.

This is where our platform stands out. We do not generate general images of cellular elements. Instead, our screening technology directly captures images of the process of mRNA translation, which shows as light pulses coming from the ribosomes as they assemble the protein. The images that look like the milky way at night tell us when, where and how much of the mRNA of the protein of interest is made in the cells in real time. We also see changes in the mRNA and how compounds are affecting its localization or modification and the effect of binding proteins.

This is far beyond what a general phenotypic screen would be able to offer since we already know that the active compounds are affecting a specific biological process around mRNA biology.

Our cloud-based AI mRNA image analysis software has already visualized and was trained on more than a billion such images and has classified them into unique “fingerprints” of mechanisms of action in mRNA biology. When an image is captured during a screen, we automatically analyze it to determine if the compound affects mRNA in a way that is similar to one of the MoA fingerprints. Following such classification, the system then suggests a small set of live biology validation experiments to further test the hypothesis, allowing us to quickly identify the MoA of the compound. With our unique technology, we were able to identify the MoAs and molecular targets of six different molecules in less than 12 months.

Read more about our mRNA translation biology