Xaira Therapeutics

Xaira Therapeutics is a drug discovery and development company applying artificial intelligence across the full span of the R&D process, from target selection and therapeutic molecule design to clinical development and patient selection. The company describes itself as building predictive and agentic AI models intended to move drug discovery from an empirical science toward an engineering discipline, with the stated aims of raising probability of success, shortening development timelines and lowering costs [0][1][3].

The company organizes its work around three elements: advanced AI research, large-scale data generation, and therapeutic product development across multiple modalities. Its computational core builds on work by co-founder David Baker and his team at the Institute for Protein Design at the University of Washington on de novo protein design, which Xaira says enables medicines against targets resistant to other approaches. A functional genomics and proteomics data platform generates and integrates multidimensional datasets spanning molecular to human scales in order to identify causal relationships in disease, and internal lab and clinical experiments are used as data-generation vehicles that feed back into model development [3].

Publicly disclosed technical output includes X-Atlas/Orion, described as the largest publicly available genome-wide Perturb-seq dataset (June 2025), and X-Cell, the company's first virtual cell model, trained on the X-Atlas/Pisces genome-wide perturbation dataset (March 2026). Company leadership has also discussed virtual cells, causal biological datasets and high-throughput biology in public podcast appearances [0][5].

Founding story

Xaira was co-founded and incubated by ARCH Venture Partners and Foresite Labs, with Nobel Laureate David Baker of the University of Washington Institute for Protein Design as a co-founder whose de novo protein design work forms the basis of Xaira's computational core [3]. ARCH's Robert Nelsen said he specifically recruited Marc Tessier-Lavigne — former Genentech chief scientific officer, Denali Therapeutics co-founder and former Stanford University president — to lead the company, and that some investors would only back Xaira if he was at the helm [6].

Business model

Xaira is developing its own internal pipeline of therapeutics across multiple modalities using its AI and data platforms, rather than describing a platform-licensing model on its public materials [3]. Sources do not describe revenue-generating commercial arrangements.

Traction

Publicly reported traction is scientific and organizational rather than commercial: release of X-Atlas/Orion, described as the largest publicly available genome-wide Perturb-seq dataset (June 2025), and the March 2026 launch of X-Cell, described as the largest virtual cell model to date and trained on the largest-ever genome-wide perturbation dataset [5]. Preprints and papers include RFdiffusion (Nature, July 2023), atomically accurate de novo design of single-domain antibodies (bioRxiv, March 2024), designed endocytosis-inducing proteins (Nature, September 2024) and the X-Cell preprint (bioRxiv, March 2026) [5]. As of March 2026 the company was still described as pre-disclosure on much of its pipeline, with an executive discussing modalities and disease areas under exploration [7].

Latest developments

In March 2026 Xaira launched X-Cell, its first virtual cell model, trained on the X-Atlas/Pisces genome-wide perturbation dataset, accompanied by a bioRxiv preprint on scaling causal perturbation prediction using diffusion language models; press coverage described it as the largest virtual cell model to date [5]. Also in March 2026, Rachel Lane was appointed SVP, Business Development and Operations, and Fierce Biotech published a podcast with President and COO Jeff Jonker on the company's strategy, modalities and disease areas [5][7]. In July 2026 Xaira announced the appointment of Ian McCaffery as SVP, Translational Science and Early Clinical Development, and promotions of Ci Chu to Chief Discovery Officer and Bo Wang to Chief AI Scientist, and published its first blog post, "Progressable Binders: The Binders that Matter" [0][5].