What happened
Xaira Therapeutics, led by Chief Discovery Officer Ci Chu and Chief AI Scientist Bo Wang, built X-Cell, an AI model trained on X-Atlas—a dataset of CRISPR-based experiments that isolate individual gene changes in human cells. The model overcame a scaling wall: earlier work on a single dataset hit a plateau at 3.1B parameters, but the new causal dataset enabled the model to continue scaling with both parameters and compute.
Why it matters
Previous RNA expression models trained on CELLxGENE (a database of 168M cells) could describe cell types and states but could not predict what happens when genes are edited or targeted by drugs—because gene expression changes are highly correlated, making causation difficult to infer. X-Cell's causal data lets the model predict real outcomes of genetic changes, potentially unlocking AI-driven drug discovery by enabling researchers to model what drugs or gene edits would do before testing them in the lab.
What to watch
Xaira's bet hinges on whether X-Cell generalizes to real lab experiments in human cells. The team abandoned autoregressive training for diffusion and reports the model beats a linear baseline that had outperformed previous models—concrete proof the approach works in validation.
Summaries like this, in your inbox every morning.
The core insight driving Xaira's work is the distinction between correlation and causation in gene expression modeling. Traditional AI models trained on CELLxGENE can capture which genes are expressed in which cells, but they cannot reliably predict what will happen when a gene is perturbed—because gene expression changes are highly interdependent, making it nearly impossible to reverse-engineer the causal direction from observational data alone. Xaira's solution is to collect experimental data where individual genes are perturbed one at a time using CRISPR, running millions of parallel tests to build a dataset that explicitly encodes causal relationships. This causal data proved to be the bottleneck: the team's earlier model plateaued at 3.1B parameters because the small, correlational dataset had exhausted its information content, but scaling the data ~30× unlocked continued improvement as parameters and compute increased—a classic signal that the model was previously starved for information, not capacity. The strategic promotion of Chu to Chief Discovery Officer and Wang to Chief AI Scientist signals that Xaira views this data-centric, causal-modeling approach as central to its long-term drug-discovery strategy.
For example, today's edition would include:
AI-summarized, only the topics you pick — one digest a day via Email, LINE, or Slack.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. Q&As are published on this page for other readers too.
Iambic and AbbVie have partnered on AI drug discovery, a collaboration reported by MobiHealthNews
Meta introduced its Muse AI agent, triggering a rally in Intel and fellow chip stocks, while AMD reached a $1T…
Qualcomm announced that its Snapdragon X Series processors will power Googlebook, a new line of laptops descri…
Anthropic has quietly set up a wet lab in the San Francisco Bay Area, its life sciences head Eric Kauderer-Abr…
Opro will hold a free online seminar on September 29 to October 1, 2026, introducing "Kamiresu," which digital…
Fujitsu said it will evolve Uvance, launched in October 2021, into an AI Transformation model, shifting to ind…
