Unlocking the Full Potential of Antibody Therapies

Episode Summary

Executive Summary: Biologic Designs CEO Yanai Ofrin argues that antibody discovery should move beyond finding binders to engineering antibodies with conditional, context-dependent functions. He explains how the company’s AI platform designs “smart” antibodies, highlights its first AI-designed antibody now in human trials, and outlines a strategy of focusing on validated immune targets while using selective partnerships to extend the platform’s reach.

Main Topics: Limitations of traditional antibody discovery (Priority: 5/5): Ofrin says standard discovery methods mostly screen for binders and rely on luck, missing antibodies’ ability to act as dynamic protein nanomachines with controllable functions. AI-driven design of conditional antibodies (Priority: 5/5): Biologic Designs uses machine learning trained on large antibody datasets to identify sequences and experimental paths that yield antibodies with desired properties such as stability, low immunogenicity, and tissue-specific activity. “Smart” antibodies and immune-system logic (Priority: 5/5): The company aims to create antibodies that can behave differently depending on biological context, such as switching between pro- and anti-inflammatory roles or acting as if-then molecular devices. Focus on validated immune targets (Priority: 4/5): Biologic prioritizes immune-related diseases and known pathways to improve existing biology with better technology, seeking wider therapeutic windows and lower toxicity rather than chasing only novel biology. Partnership strategy and value creation (Priority: 4/5): Ofrin describes a selective partnering model aimed at strategic deals with upfront non-dilutive cash, senior-management support, and upside, while keeping much of the company focused on its own pipeline. IL-2 program and Aulis Bioscience (Priority: 5/5): The company’s collaboration with Apple Tree Partners created Aulis to advance AU-007, described as the first computationally designed antibody to enter human trials, aimed at modulating endogenous IL-2 and potentially avoiding negative feedback seen with engineered IL-2 approaches. Capital efficiency and financing model (Priority: 3/5): Biologic says it has built a sizable organization and multiple partnerships with relatively little dilutive capital, relying heavily on non-dilutive deal cash while still seeking additional funding.

Key Arguments: Traditional antibody discovery is optimized for binding, not for controlling function; this leaves major therapeutic potential untapped. Antibodies can be designed to act conditionally, changing behavior by tissue, metabolite, or immune context. AI can help specify not just final antibody sequences but also the experiments needed to generate better training data for a project-specific model. Validated immune pathways are attractive because smarter antibodies may improve efficacy and safety where older drugs have plateaued. Partnering should be selective and strategic; otherwise value creation can be delayed and diluted by long pharma timelines. Endogenous cytokines, especially IL-2, should be controlled directly rather than replaced with engineered versions that ignore feedback loops. The company’s platform can reach previously “undruggable” targets by designing exact functional interactions instead of screening random binders. Aulis and AU-007 serve as key external validation that AI-designed antibodies can reach the clinic and behave like real drugs.

Data Points: Company age: About 10 years old - Biologic Designs has grown from a university lab into a company over roughly a decade. Headcount: About 60 people - The company now has employees in Israel and Boston. Dilutive investment raised: Less than $30 million - Ofrin emphasized capital efficiency despite clinical-stage progress and multiple partnerships. Non-dilutive cash: Not quantified - He said partnerships supplied substantial non-dilutive funding that helped support operations and programs. AI-designed antibody in clinic: First computationally designed antibody entering human trials - Ofrin said AU-007 is the first AI/computationally designed antibody dosed in humans. Partnership year: 2021 - Two collaborations discussed—Nectar Therapeutics and Lilly—were announced in 2021. Conference mention: 17th Annual Outsourcing and Clinical Trials New England 2025 Conference - Included in sponsor copy at the beginning of the episode. Conference dates: October 15th and 16th, 2025 - Sponsor promotion for Praveil Infoworks event appearance. Lab experiment scale: About a billion different experiments - Ofrin described the AI system outputting a very large experimental program to generate training data. Disease focus: Autoimmune diseases and immuno-oncology - These were described as the core areas for Biologic’s own pipeline.

Pivotal Quotes: "Antibodies are protein. Proteins are clever nanomachines. They can execute sophisticated programs that are encoded in them." — Yanai Ofrin: He explains why Biologic thinks antibodies should be engineered for function, not just binding. "Instead of looking for new biology and attacking it with old technology, we are focusing on old biology and attacking it with new technology." — Yanai Ofrin: He describes Biologic’s core strategy of using AI-designed antibodies against validated pathways. "We are basically hijacking the communication between immune cells and making sure that every time immune cells want to communicate with each other, we are interfering in this message." — Yanai Ofrin: He explains how the AU-007/IL-2 approach may modulate immune signaling in cancer.

Implications: If validated clinically, Biologic’s approach could shift antibody discovery from binder screening to function design, improving drugs for validated immune targets and enabling context-sensitive therapies that are more effective and less toxic.