Living, full-thickness human skin, that stays viable outside the body for a month or more – far longer than conventional skin explants – forms the data foundation for AI-driven discovery of new skin bioactives
The validation is public: a bioRxiv preprint documents Yuna reproducing three distinct skin biologies across the four-week window
Outer Bio, a discovery company for skin bioactives, today emerged from stealth and released Yuna, an ex vivo human-skin platform that keeps full-thickness skin alive and measurable for over four weeks — substantially longer than standard ex vivo skin — with its native architecture and immune function intact. Yuna is the data-generation and validation engine beneath Outer Bio’s approach: a way to generate reproducible, high-fidelity, clinically relevant human-skin data at a duration and level of detail the field has lacked, and the foundation for the company’s machine-learning-driven discovery of new bioactives.
The skincare industry generates more than $100 billion a year, yet in three decades it has added essentially one FDA-approved topical mechanism that meaningfully improves visible skin aging: retinoids. The constraint has been technical, not commercial. Taken out of the body, human skin survives in the lab for less than a week. But the processes that determine whether and how a compound works — collagen remodeling, chronic inflammation, cellular senescence — unfold over weeks. The field has been forced to judge long-term biology inside a short-term window.
Yuna closes that gap. It holds full-thickness human skin including epidermis, dermis, multiple cell types, and resident immune function alive for four or more weeks. The tissue is sourced from donors across ages, sexes, and Fitzpatrick skin types I through VI and read out by longitudinal multi-omic measurements spanning transcriptomics, histology, and secreted proteins, building an ever-expanding dataset that improves the platform’s performance over time. Each compound-sample pair generates more than 30,000 data points; to date, the platform has produced over 10 terabytes of proprietary data across more than 300 donors and 10,000 treatments. Because the platform relies on living human tissue rather than a reconstructed model or a cell line, signals observed in Yuna are grounded in native human skin biology and can support clinical translation.
On that foundation, Outer Bio is building an integrated discovery engine that pairs machine learning with every experiment: models trained on the platform’s data prioritize which compounds to test, and each result sharpens the next round of predictions. The company’s virtual screening library already spans 5.9 million compounds across 11 databases, with a roadmap toward billions of synthesizable structures.
“Skincare has never lacked demand or ambition. It has lacked the ability to study compounds and mechanisms in living human skin over a biologically relevant timeframe,” said Michael Polansky, co-founder and CEO of Outer Bio. “Most experimental platforms rely on reconstructed skin or cell monolayers. Those models can tell you whether a compound is toxic; they can’t tell you whether it works, because they lack the complexity and architecture of the real organ. And they fail within days, before the biology that matters even begins. We built Yuna to close that gap.”
“AI for biology has a fundamental data problem,” Polansky added. “AI models are improving far faster than the experiments that validate and train them. The bottleneck isn’t intelligence—it’s a scalable source of clinically valid data, and that data has to be grown, not computed. Yuna is built to generate it.”
The platform’s validation is now public. In a preprint posted to bioRxiv, Outer Bio documented Yuna reproducing three distinct skin biologies across a four-week window: a psoriasis-like inflammation sustained for three weeks and reversed by a clinically used JAK inhibitor; senescence signatures in aged donor tissue reduced by a senolytic combination; and a UV-induced photoaging injury mitigated by a topical sunscreen. Each is a result that conventional simplified or short-duration models cannot fully reproduce, because the relevant biology depends on both tissue complexity and time. Beyond the preprint, Outer Bio has tested and validated Yuna on a broad set of benchmark bioactives and therapeutic agents across anti-aging, antioxidant, anti-inflammatory, senolytic, and skin-protective applications, building an internal reference dataset that strengthens the platform over time. The same ability to reproduce known compounds is what gives Yuna the foundation to discover new ones.
While Yuna is built primarily for internal discovery, Outer Bio is making it available through select partnerships. Biopharma teams, ingredient suppliers, and consumer brands can access the platform for long-duration efficacy studies, skin-toxicity assessment, and collaborative discovery programs targeting specific skin-biology applications.
About Outer Bio
Outer Bio is a discovery company for skin bioactives, the compounds that make skincare work. The company generates high-fidelity, clinically relevant human-skin data at a scale and duration the field has lacked, and uses it to discover novel bioactives across consumer skincare and therapeutics. Its core platform, Yuna, keeps living, full-thickness human skin viable for four or more weeks; the validation is documented in a preprint on bioRxiv. Outer Bio is backed by Wing Venture Capital, Initialized Capital, Sozo Ventures, Hawktail, Lightspeed Venture Partners, Third Kind Venture Capital, Liquid 2 Ventures, and SV Angel. For more information, visit outerbio.com.
View source version on businesswire.com: https://www.businesswire.com/news/home/20260820678439/en/
