Here’s a stat that should bother you: one-third of all FDA-approved drugs target G-protein coupled receptors. GPCRs are arguably the most validated drug target class in all of pharmacology. And yet, in the antibody space — where the industry has poured hundreds of billions of dollars — almost no GPCR-targeting antibodies exist. Roughly 220 disease-linked GPCRs remain completely undrugged by antibodies. The reason? Traditional antibody discovery methods just can’t handle them.

Antiverse thinks AI can fix that. The Cardiff, UK-based company just closed a $9.3 million Series A led by Soulmates Ventures, bringing total funding to over $20 million. Proceeds are going toward expanding the company’s AI-driven de novo antibody design platform and advancing internal pipeline programs toward in vivo efficacy studies.

The platform uses machine learning to model antibody-antigen interactions and generate target-specific libraries from structural and sequence data — essentially designing antibodies computationally before ever touching a wet lab. In a collaboration with a top-20 pharma company, the platform identified 248 sequences, with over 230 confirmed as actual binders. That’s a hit rate that traditional phage display can’t touch. And they’re doing it in roughly six months per target.

The validation signals keep stacking. Alongside the Series A, Antiverse announced a research agreement with the Cystic Fibrosis Foundation focused on using the platform against GPCRs and ion channels relevant to CF. They also have a multi-target partnership with Nxera Pharma (formerly Sosei Heptares) — combining Antiverse’s generative AI design with Nxera’s NxWave platform for GPCR structural determination. Two collaborations with top-20 pharma companies. And they’ve already generated functional antibodies against PAR1 and anti-PD-1 candidates progressing to preclinical development.

This is the kind of AI-in-biotech story that actually holds up under scrutiny. Not “we have a model” — but “we have a model, it produces confirmed functional binders at scale, and disease foundations and top-20 pharma are paying to use it.” The undruggable proteome isn’t going to drug itself. Antiverse is making the case that AI can unlock the targets traditional methods can’t reach.