Seamless integration of data layers across preclinical drug discovery.
High-resolution chemical fingerprinting of thousands of biosamples using metabolomics
Generative deep-learning to identify chemical structure and properties of novel chemistry at scale
Rapid pooled characterization of biological activity across target-binding, phenotypic, and protein-protein interaction assays
Library-wide, untargeted characterization of oral bioavailability and tissue distribution
We are creating a map of the world’s chemistry by profiling 1000s of complex samples through tandem mass-spec
We created generative deep learning models to predict chemical properties and structure directly with high accuracy from MS2 spectra, allowing us to characterize novel chemical space orders of magnitude faster than previously possible.
We have developed proprietary high-throughput laboratory methods to link each compound to laboratory-tested biological activities, finding not one, but hundreds of active compounds at once.
Our bioactives bind undruggable targets, inhibit high-value pathways using novel mechanisms, and modulate PPIs across the proteome.
We have invented methods to obtain organ distribution data in vivo for 1000s of compounds at once with high-confidence.
Fueled by the unique challenges of working with mixtures of unknown and highly-diverse compounds, we redesigned the early-stage drug discovery process.
Instead of screening for one kind of activity and then repeating the process, we annotate our entire library across dozens of bioactivity assays and organ distribution experiments. Our platform in action: