{"slug": "nabla-bio-expects-human-trials-of-ai-designed-antibody-drugs-within-two-years", "title": "Nabla Bio expects human trials of AI-designed antibody drugs within two years", "summary": "Nabla Bio, an OpenAI-backed Massachusetts biotech company, said its JAM-2 model generated high-affinity antibody candidates against 26 targets and that it expects first-in-human trials of AI-designed antibodies within one to two years. More than 50% of the candidates met developability criteria without optimization, and JAM-2 reached up to an 11% success rate for direct on-cell binding against GPCRs, the target class for roughly a third of FDA-approved drugs. The timeline follows Nabla's October 2025 Takeda partnership expansion, which included double-digit-millions upfront payments and potential milestones exceeding $1B.", "body_md": "# Nabla Bio expects human trials of AI-designed antibody drugs within two years\n\nThe biotech startup's JAM-2 model generated high-affinity antibody candidates against 26 targets, with some matching or surpassing traditional therapies in binding strength.\n\nNabla Bio, a Massachusetts-based biotech company backed by [OpenAI](https://cryptobriefing.com/markets/openai/), is preparing to take AI-designed antibodies from the lab into human bodies. The company’s latest model, JAM-2, has produced antibody candidates with binding affinities that match or beat conventional therapies, putting it on track for first-in-human trials within the next year or two.\n\n## What JAM-2 actually does\n\nNabla’s JAM-2 model generated antibody candidates against 26 different targets. For nearly half of those targets, the candidates achieved picomolar to single-digit nanomolar affinities. More than 50% of the candidates met developability criteria without any optimization.\n\nJAM-2 achieved up to an 11% success rate for direct on-cell binding against GPCRs. GPCRs are the target class for roughly a third of all FDA-approved drugs, yet designing new antibodies against them has historically been a grind.\n\nSome of the AI-generated candidates demonstrated the ability to activate cellular signaling pathways, a capability that opens the door to agonist therapies where the drug needs to turn something on, not just block it.\n\nThe model also showed strong epitope precision, routinely hitting 30-70% of user-defined epitopes, meaning researchers can specify exactly where on a protein they want the antibody to bind.\n\n### AI, tech, and the markets they move—in one daily briefing.\n\nDaily. Free. Join 34,000+ readers across crypto, finance, and policy.\n\nJAM-2 employs test-time scaling, an approach inspired by the reasoning methods used in systems like OpenAI’s ChatGPT, where the model iterates through multiple rounds of reasoning to refine its outputs rather than generating an answer in one pass.\n\n## The Takeda connection and commercial momentum\n\nNabla expanded its partnership with Takeda in a deal announced in October 2025. The terms included upfront payments in the double-digit millions, with potential milestone payments exceeding $1B. Following that partnership expansion, Nabla indicated it expected to generate first-in-human data from its AI-designed molecules within one to two years.\n\n## A crowded but early field\n\nNabla isn’t the only OpenAI-backed company working on AI-driven antibody design. Chai Discovery raised $400M in a Series C funding round at a $3.8B valuation in mid-2026 but has not confirmed clinical assets or trial timelines.\n\n## What to watch\n\nInvestors watching the AI-biotech space should focus on three things: whether Nabla files an Investigational New Drug application within its stated timeline, how the Takeda partnership molecules perform relative to traditionally developed candidates, and whether Chai Discovery or other competitors announce their own clinical milestones.\n\nTakeda’s deal structure alone implies that successful development could generate over $1B in payments to Nabla.\n\n**Disclosure:** This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our\n\n[Editorial Policy](https://cryptobriefing.com/editorial-policy/).", "url": "https://wpnews.pro/news/nabla-bio-expects-human-trials-of-ai-designed-antibody-drugs-within-two-years", "canonical_source": "https://cryptobriefing.com/nabla-bio-ai-antibody-human-trials/", "published_at": "2026-09-28 04:22:49+00:00", "updated_at": "2026-09-28 04:48:43.333182+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-startups", "ai-research"], "entities": ["Nabla Bio", "JAM-2", "OpenAI", "Takeda", "Chai Discovery", "GPCRs", "FDA", "Diego Almada Lopez"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/nabla-bio-expects-human-trials-of-ai-designed-antibody-drugs-within-two-years", "markdown": "https://wpnews.pro/news/nabla-bio-expects-human-trials-of-ai-designed-antibody-drugs-within-two-years.md", "text": "https://wpnews.pro/news/nabla-bio-expects-human-trials-of-ai-designed-antibody-drugs-within-two-years.txt", "jsonld": "https://wpnews.pro/news/nabla-bio-expects-human-trials-of-ai-designed-antibody-drugs-within-two-years.jsonld"}}