Introducing onepot CORE 2 Triangulum
Seven months ago, we released the first version of onepot CORE, our chemical compound space. Since then, we have worked with dozens of partners to make the CORE larger, more chemically diverse, and more synthetically reliable. Every order and every experiment taught us something: which regions of chemical space were most useful, where synthesis failed, which protocols needed improvement, and which new reactions were worth bringing onto the platform.
Today, we are releasing onepot CORE 2 Triangulum, a chemical dataset containing 30 billion compounds across 16 unique transformations.
onepot CORE has been learning
Over the last seven months, we have continuously expanded the chemistry behind the CORE and improved the experimental protocols used to make its compounds. In the lab, our AI scientist Phil has driven much of the discovery work behind new reactions and reaction conditions, working alongside our scientists and robotic infrastructure.
At the same time, we have been rebuilding the CORE for computational workflows. Enumerating billions of molecules is relatively easy; searching them efficiently is not. We developed generative models and a proprietary synthon-based approximation of our ML filters that allow computational systems to navigate much larger compound spaces without evaluating every individual molecule independently.
Introducing Triangulum
CORE 2 expands the synthetically accessible onepot compound space to 30 billion compounds derived from 16 transformations, including multicomponent reactions and multistep chemistry. Triangulum also pushes significantly further beyond conventional lead-like chemical space, introducing more complex molecules.
Some of the notable compound classes now represented in the CORE:
Including acrylamides and other targeted covalent warheads.
Multifunctional probes for target identification and chemoproteomics.
Cubanes, bicyclopentanes, and other fused and caged systems.
Building blocks for targeted protein degraders and molecular glues.
Larger, more complex molecules aimed at extended binding interfaces.
A chemical space connected to a lab
Every compound in the CORE is connected directly to onepot's synthesis infrastructure. You can search or design molecules computationally, submit them through onepot, and receive physical material — without designing the synthesis route yourself or coordinating a traditional outsourcing workflow.
Simple pricing
We have also simplified CORE pricing. Compounds are now available in two tiers, depending on the chemistry required.
Per compound, for targets that run on our established transformations.
Per compound, where the route calls for more demanding chemistry.
As always, you do not pay for failed experiments. Our standard target is at least 0.5 mg of material at ≥90% purity. If a compound falls below those specifications, we can either provide it at a discount or not ship it at all — your choice.
Built for computational drug discovery
CORE 2 Triangulum was designed not only for medicinal chemists selecting compounds individually, but for computational systems selecting hundreds, thousands, or eventually millions of candidates.
The CORE can be integrated directly into computational workflows through the onepot API, allowing models and agents to move from molecule design to physical synthesis programmatically. This interface between computation and the physical lab is becoming increasingly important with the adoption of AI systems for drug discovery and agrochemistry research.
CORE 2 Triangulum is available today
Request access to the space, reach out to your contact at onepot, or start using the CORE compound space through the API.
For anything else, including how Triangulum fits into your existing screening or generative workflows, reach out to hello@onepot.ai .