San Diego sequencing heavyweight Illumina on Tuesday pulled back the curtain on the Billion Cell Atlas, a massive dataset that tracks how one billion individual cells respond when genes are switched on or off. The company says it is built to train AI systems to spot drug targets and speed early-stage discovery. It is the first installment in a planned multi-billion-cell program that zeroes in on more than 200 disease-relevant cell types. In a city already synonymous with genomics tools and scrappy biotech startups, the atlas is Illumina’s bid to become a go-to supplier of foundational biological data for pharma and AI teams.
In a company release, Illumina said the Billion Cell Atlas will log how one billion single cells respond to CRISPR-based genetic perturbations, with AstraZeneca, Merck and Eli Lilly signed on as founding participants. "We believe the Cell Atlas is a key development that will enable us to significantly scale AI for drug discovery," Illumina CEO Jacob Thaysen said in the release.
How the atlas measures biology
The dataset captures transcriptomic changes at single-cell resolution after targeted gene edits, letting researchers see which genes are driving disease pathways and drug responses, according to Reuters. By tracking how each cell reacts when specific genes are turned on or off, scientists can train models that predict mechanisms and test candidate targets faster than running many small, bespoke lab assays.
Platform and scale
Illumina says the project runs on its Single Cell 3' RNA prep platform, processes data through the DRAGEN pipeline, and hosts results in Illumina Connected Analytics. That setup can capture millions of cells per experiment and generate petabytes of single-cell transcriptomic data each year, according to Illumina. The company is positioning the Billion Cell Atlas as the debut product from its new BioInsight business, which couples large standardized perturbation datasets with algorithms for drug discovery and target validation.
Big pharma signs on - quickly
Local coverage says the rollout got attention almost instantly: The San Diego Union-Tribune reports that additional pharmaceutical and tech firms reached out to Illumina within hours of the announcement. With founding partners already at the table, Illumina plans to offer the atlas as a resource for target validation, virtual cell modeling and AI training workflows used by its pharma collaborators.
Why it matters
The move plugs directly into a broader push to ground AI-driven drug discovery in hard biological data rather than relying mostly on literature mining or trial-and-error screening. Reuters notes that companies such as Certara, Schrodinger and Recursion are already building computational models to predict absorption, distribution and toxicity. Growing regulatory interest in alternatives to animal testing is also steering developers toward high-fidelity datasets that can support predictive models.
Money and timing
Illumina has been reshaping its business mix even as its financial picture shifts. The company reported third-quarter revenue of about $1.08 billion and GAAP net income of $150 million, down from $642 million a year earlier, according to the Associated Press. Coverage of Illumina's investor materials indicates the company has released preliminary fourth-quarter numbers and said it will report full fourth-quarter and fiscal-year 2025 results after the market closes on Feb. 5, per reporting on the company's investor statements.
For San Diego’s life-science ecosystem, the Billion Cell Atlas is both bragging rights and potential fuel. Local labs, startups and big pharma players could all tap standardized perturbation data to move programs faster and trim early-stage failures. Whether that mountain of data actually translates into quicker new medicines for patients will hinge on how accessible and interoperable Illumina makes the resource for outside researchers.
