Creating the blueprint
for medicines on demand.

Better and safer drugs faster, by design.

Creyon Bio is taking an engineering approach to creating new oligonucleotide-based medicines (OBMs). We bring together insights from modern biology, genomics, pharmacology, biophysics, and chemistry with the speed and predictive power of machine learning and AI. Creyon’s advanced machine learning and AI-backed platform maps sequence and foundational biophysical properties of existing and novel OBM chemistry to build accurate predictive models of safety and efficacy. The platform works by identifying the design rules and engineering principles for OBMs — including modalities ranging from single-stranded antisense oligonucleotides (ASOs) that reduce gene expression levels or change splicing events to small interfering RNA (siRNA), to DNA and RNA editing systems, to even targeting aptamers. By engineering OBMs with predictable safety and performance before they’re even tested, we are enabling the vision of on-demand, gene-centric therapeutics for all and making prolonged and expensive drug discovery, as we know it, a thing of the past.

Modular building blocks for precision-engineered drugs.
A scientific illustration showing RNA from transcription in the nucleus to translation in the cytoplasm and how OBMs can regulate mRNA at every step. A small robot hand icon marks the steps.

OBMs are an ideal modality for creating gene-centric therapies. Unlike other drug modalities such as small molecules or antibodies, a principal way OBMs interact with targets is through well understood Watson-Crick-Franklin hybridization: a zipper-like process that selectively and precisely targets specific positions of DNA or RNA molecules within the cell. They can control the effective activity of genes through multiple mechanisms. They can enhance or block molecular interactions which can control how genetic information is used by the cell.  This provides the foundation for creating new medicines which directly address the genetic and molecular basis of disease – from preventing the production of toxic RNAs or protein products, to increasing the levels of insufficiently expressed enzymes, to even editing mRNA or changing how the splicing machinery assembles the genetic instructions which control protein production. Creyon’s technologies and efforts make engineering new OBMs to treat disease through these varied mechanisms, and perhaps yet to be discovered mechanisms, not only possible, but fast and efficient. We are even engineering delivery systems to directly target OBMs to specific cell types and tissues.

A scientific schematic of single-stranded RNA is on the far left. In the middle, a zoomed-in single nucleotide shows a hexagonal base, a pentagonal base, and two linkers. A collection of buckets below each piece show a vast multitude of linker, sugar, and nucleobase variations receding into the distance.

But the sequence of an OBM is just one aspect of what makes it work. From the sequence composition itself to the sugar, nucleobase, and backbone chemistries, every component is modifiable and can have critical implications for its pharmacology, from its safety and efficacy to how long its effects will last. Yet there are 4 billion possible sequence configurations in a molecule that is just 16 bases long. Add a few sugar and backbone modifications, and that number skyrockets to more than billions of possible designs — an impossibly large number of possibilities that could never be screened or explored with traditional methods. The Creyon™ Platform creates the right data to explore this space far more efficiently and develop predictive models that integrate learnings across hundreds of thousands of data points to create new medicines that are knowably safe and effective. 

Creyon’s groundbreaking approach is upending the process of drug development by uncovering the engineering principles that allow us to employ the right designs right away. With a deliberate approach to engineering medicines instead of trial-and-error screens, Creyon is engineering molecules that can be precisely tuned with predictable results before they’re even tested.

Purposeful data for powerfully accurate models.

Creyon’s work is powered by unique datasets that have been optimally engineered for active machine learning and yield unprecedented knowledge orders of magnitude more efficiently than current methods.

Building on these datasets, Creyon’s machine learning models generate powerful knowledge about the biophysics and mechanisms that impact how OBMs work. The models uncover the chemical design rules and engineering principles that allow us to create optimized OBMs from scratch.

Trained on our bespoke datasets, our models generate not only accurate predictions about safety and efficacy, but also actionable knowledge that connects the emergent properties of the chemistry and sequence of an OBM with its pharmacology. In this way, Creyon can explore novel OBM chemistries with unparalleled efficiency. Further, we can focus our methods to optimize OBMs for any use, including knockdown, controlling splicing, editing, enhancing or inhibiting regulatory interactions with RNA.

Learn how we’re building a world with no gap between diagnosis and treatment, unconstrained by disease incidence or distribution.

Our Work