Skip to content
Drug Discovery

RubrYc Discovery Engine

Using Artificial Intelligence to Discover unique hard-to-find antibodies

With only five of every 5,000 concepts making it into the clinic1, we think there’s a better way.

That’s why we partnered with RubrYc Therapeutics in 2021, licensing a particularly difficult-to-design immunotherapy candidate and why we’re extremely excited we’ve now acquired the RubrYc Discovery Engine. The announcement of our acquisition can be found here.

We expect RubrYc’s artificial intelligence (AI)-Driven Drug Discovery Engine to significantly strengthen our drug discovery capabilities as an integral part of our early stage drug discovery center in San Diego, California.

So, what’s so novel about another “AI” Platform?

In short, the AI Drug Discovery Engine is designed to find antibodies against the most challenging targets – with the greatest potential success rates – where others have struggled or failed.

It does this by identifying, engineering, and optimizing artificial proteins that mimic the exact shape of complex epitopes, including subdominant epitopes, to discover highly promising therapeutic antibodies. An epitope is the specific section of a target protein to which antibodies bind. The range of potential epitopes or binding sites on a target is enormous. But not all epitopes are created equally.

Traditional drug discovery has allowed the immune system to do the picking of epitopes and thus biasing antibody selection towards the ones most commonly recognized by the immune system. Often, the best epitope for developing a new drug is harder to find or is overshadowed by other more immunodominant epitopes.

That’s where the AI Discovery Engine comes in.

The AI Discovery Engine uses proprietary computational biology, predictive algorithms and 3-D modeling for the identification and engineering of proteins that mimic epitopes that have proven difficult to target using standard immunization and screening strategies.

Combining Computational Biology and 3D-Modeling for Identifying & Engineering Large-Molecule Drug Candidates

Step 1. Epitope Targeting Engine: Unparalleled AI-powered precision targeting hard to target epitopes. Step 2. RubrYcHu™ Library: AI-generated human library, free of sequence liabilities provides a unique pool of mAbs. Step 3. Human Antibody Hits: Epitope-specific mAb panels are rapidly processed through screening funnels. Step 4. Optimize with StableHu™ AI Engine: AI-powered sequence optimization to improve performance. Step 5. Optimized mAb Lead Pool: Optimized Leads are evaluated and ranked in translational disease models. Final: On-Epitope Clinical Candidate.

We like to think of it as hooking the big catch. Just as fishermen carefully select their bait to attract the right fish – a common worm might only attract a common fish – we use highly selective (not common) epitope structures to lure the rarest and most promising of antibodies.

After licensing the first antibody from RubrYc, now IBIO-101, we were able to advance the molecule into the IND-enabling stage in less than 12 months, due in part to its high-quality design. Along with acquiring the RubrYc Discovery Engine, we’re adding four of RubrYc’s pre-clinical programs to our pipeline.

AI-Powered Generation and Evolution of Stable Artificial Epitopes as ‘Lures’ for Target Epitope-Specific Antibodies

Artificial epitope evolution diagram shows target epitope becoming a stable artificial epitope as antibody 'lure'

  1. Devol, Ross & Bedroussian, Armen & Yeo, Benjamin. (2022). The Global Biomedical Industry: Preserving U.S. Leadership.