Xaira Therapeutics

Research on the record

• 2026-10-23Xaira built a camera-guided arm for instruments made for human hands, and over seven days it loaded 18 plates into an HPLC, 80% more than its model of weekday hoursLevel 4 · Tested in the physical world
• 2026-10-21Xaira's X-Scientist agents picked genes to perturb in liver cells, and 7.1% of its picks reversed the TGF-beta signal against 6.9% for its own scientists and 0.9% at randomLevel 4 · Tested in the physical world
• 2026-10-02Xaira says X-Design Vega took its XA-1 nanobody from DNA synthesis to a lead in 7 weeks, and made an antagonist for a GPCR that library screening and llama immunisation failed onLevel 3 · Checked against existing evidence
• 2026-03-18Xaira's X-Cell, trained on 25.6 million perturbed cells, beats earlier virtual-cell models up to five-fold on its own held-out genes, but its test loss barely improves with scaleLevel 2 · Computed
• 2025-06-16Xaira releases X-Atlas/Orion, a genome-wide Perturb-seq atlas of eight million cells, with the platform that made itLevel 1 · Exists

Graded by how far the evidence went, not how good the result is. See the methodology.

Read on AI4S.fyi

• 2026-10-12Ten years, tens of millions of cellsCommentaryNature Biotechnology editorial · Nature Biotechnology
• 2026-05-21Building the Drug Discovery Engine of the Future with AI-Empowered Nodal BiologyCommentaryAnna Greka · Dædalus

Founders' research on the record

Papers and results a founder of the company authored, from the timeline's bylines. Most come from before the company existed: they are the team's track record, not results the company has reported.

• 2026-10-26A designed protein scaffold binds a metal photocatalyst, and five evolved mutations turn it into an enantioselective enzyme reaching 50,000 turnoversLevel 4 · Tested in the physical world
• 2026-10-21A single protein-design model from Baker's lab and NVIDIA beats the best specialist tools on four design tasks, and its designs were made and measured in the labLevel 4 · Tested in the physical world
• 2026-10-14Designed from scratch, 13 of 69 cysteine proteases cut the peptide backbone where told, and designs cut their targets in cellsLevel 4 · Tested in the physical world
• 2026-10-14Zinc proteases designed with RFdiffusion cut chosen peptide bonds, and some cut TDP-43 and amyloid-betaLevel 4 · Tested in the physical world
• 2026-10-02Baker lab designs two working enzymes with complex active sites, one with a 2.7 Å crystal structureLevel 4 · Tested in the physical world

Graded by how far the evidence went, not how good the result is. See the methodology.

People

Only people the AI4S.fyi record has named: in a round, a timeline entry, commentary, an interview or a sourced role. Not the full team.

Funding rounds tracked

• 2024-04-23Round (unnamed) · more than $1 billion in committed capitalconfirmedLed by ARCH Venture Partners, Foresite Capital · With F-Prime, NEA, Sequoia Capital, Lux Capital, Lightspeed Venture Partners, Menlo Ventures, Two Sigma Ventures, Parker Institute for Cancer Immunotherapy, Byers Capital, Rsquared, SV AngelStated use: build the AI research, data generation and drug development platform

Each round links to its line on the funding record, with its source.

Also in Drug discovery on the funding board

The other companies the funding board files under the same field, latest round first. A list of what the record tracks, not a ranking or a market map.

• 2026-10-29ALLOXSeed · US$10.5M (€9.4M)Measures the effects of thousands of protein mutations at once to map allosteric and other regulatory sites, and builds AI models on that data to find drug-binding spots on hard targets such as protein-protein interactions and transcription factors, for its own small molecules.
• 2026-10-27Atlas BioSeed · $10MRuns in silico clinical trials, starting in immunology: a foundation model trained on about 2 million RNA readouts from human tissue predicts each virtual patient's response to several treatments and diseases, to help drug developers choose what to take into Phase II.
• 2026-10-23Pathos AISeries E · $237MLicenses clinical-stage drugs from other companies and develops them itself, using an agent platform (Foundry, earlier PathOS) trained on patient records to choose assets, pick the patients most likely to respond and adjust trials as they run. Its pipeline is in cancer: prostate cancer, multiple myeloma, lung and breast cancer and TROP2/HER3-expressing solid tumours; in October 2026 it began work in Alzheimer's disease with Bristol Myers Squibb and Tempus.
• 2026-10-23Zhichong PharmaSeries A1 · tens of millions of yuanDevelops new drugs for pets with generative AI, using its own small-molecule and large-molecule design engines (ChemoFetch and OmniFetch), and argues that testing in the target animal closes the loop from design to evidence faster than human drug development can; its programmes are in anti-infectives, chronic disease and ageing.
• 2026-10-14Gaowei YiyaoPre-A · undisclosed; PEdaily says the company has raised several hundred million yuan so far this yearMines protein-interaction and other high-dimensional biology data with its own models to find drug targets others treat as undruggable, then runs its own small-molecule and antibody programmes. Its GROOVY platform found the target behind GW-01, a Notch-pathway inhibitor now in Phase I; ACUMEN maps pairs of neighbouring cell-surface proteins for bispecifics and NEXUS looks for tumour neoantigens.
• 2026-10-14Iambic TherapeuticsIPO · $150MDesigns small-molecule cancer drugs with AI structure-prediction models (NeuralPLexer) linked to an automated chemistry and biology lab.

All 37 in this field →