A small Bay Area biotech is asking antibodies to do more than stick. The ask: switch the lock, don't just hold the key.
A 23-person team. A yeast colony with a job. A wall of pipettes lit by a screen showing GPCR activation curves. The pour-over is still warm - the CEO doubles as Head Barista.
Binding is a handshake. Function is a conversation. The pharmaceutical industry has spent forty years getting very good at handshakes - measuring whether an antibody and a target stick to one another. Abalone Bio, founded in 2018 in a corner of MBC Biolabs, has built a platform around a different question: when the antibody arrives, does the target actually do something?
It is a deceptively simple shift. It is also, depending on whom you ask, either the future of antibody drug discovery or a very expensive experiment in yeast genetics. Richard Yu, the company's co-founder and chief executive (and, per his LinkedIn, Head Barista), prefers the first framing. The peer-reviewed paper his team published on SARS-CoV-2 neutralizing antibodies suggests he might be onto something.
FAST stands for Functional Antibody Selection Technology. The mechanism is biology used as a scoreboard: engineered yeast cells are designed so that their survival depends on whether a given antibody activates the target. Live cells mean the antibody worked. Dead cells mean it didn't. Hundreds of millions of antibodies are screened in parallel - roughly a hundred times more than competing approaches.
The readout is biological activity. The platform answers what matters in a drug.
Functional datasets feed protein language models that propose better next-round antibodies.
GPCRs and other historically difficult targets where small molecules struggle.
Each experiment trains the model. Each model proposes the next experiment.
Approximate, per company-disclosed figures.
G-protein coupled receptors are the workhorse of pharmacology. Roughly one in three approved drugs targets a GPCR. The catch: most of those drugs are small molecules. Antibodies, with their precision and long half-lives, would be useful here - but discovering antibodies that turn on a receptor, rather than block it, has been a structural problem for decades.
Abalone Bio's pitch is that yeast biology can break the bottleneck. If the platform delivers, the implications extend to obesity, metabolic disease, and a long list of conditions where the existing drugs are either blunt instruments or come with side effects nobody loves. The team has already disclosed activator antibodies for two GPCRs - early, but unusual.
Co-founders Richard Yu and Gustavo Pesce met in 2003 at the Molecular Sciences Institute, where they worked on systems and synthetic biology. They co-founded an algae biofuels startup. They went separate ways through Bolt Threads and Amyris. In 2018, they regrouped around a different question: what if the discipline they'd built in industrial synbio could be pointed at therapeutics?
Co-Founder · CEO · Head Barista
Former Scientific & Operations Director at MBC Biolabs and Principal at Mission Bay Capital. Previously an investigator at The Molecular Sciences Institute studying how cells process information.
Co-Founder
Synthetic biology veteran. Collaborator with Yu across four ventures spanning algae fuel, industrial materials, and now antibody therapeutics.
Chief Scientific Officer
Leads the scientific direction of Abalone Bio's antibody discovery and engineering programs.
VP, Data Science & AI
Runs the machine learning side of the platform - turning functional datasets into next-round antibody designs.
Things you notice quickly about Abalone Bio: the technical density of the founding team, the absence of corporate jargon on the company website, and the fact that the CEO's LinkedIn title includes the words Head Barista. These are not accidents. The company has stayed small (23 people), kept a tight focus on a single platform, and stretched its capital - $4.46M in equity has been supplemented by roughly $7M in non-dilutive grants. For a preclinical biotech, that is unusual hygiene.
The pour-over is cold now. The plates from last night have been read. Somewhere in the data, a handful of yeast colonies survived because the antibodies they carried did exactly what the team hoped: switched a receptor on. The Head Barista is on his second cup. The 23-person company has not yet cured anything. But they've changed what the question is, and the question is the part that compounds.
Binding was the handshake. Function is the conversation. Abalone Bio is the one trying to teach antibodies to talk back.