Problem

Blank Bio targets a central bottleneck in precision oncology: clinical trials often average outcomes across responders and non-responders, allowing non-responders to dilute the measured drug effect and mask benefits for the patients who respond. Standard bulk RNA-seq workflows further compress tumor data into per-gene counts, discarding isoform architecture, mutational complexity, and other patient-specific biology that could improve clinical interpretation.

The economic pain is failed or underpowered drug development: the company says biomarker-enriched trials have helped distinguish billion-dollar blockbusters from failed drugs in the same class, while covariate-adjusted prognostic scores could help trials reach statistical power with fewer patients. Its clearest use case is therefore helping pharmaceutical teams identify likely responders and select the right patients for a trial, while also predicting how each patient's disease would progress without treatment.

Product / Service

Blank Bio is an applied AI research lab that trains RNA foundation models on bulk RNA-seq and related transcriptomic data to learn patient heterogeneity. The models are designed to answer two trial-critical questions—how a patient's disease is likely to evolve and who is likely to respond to a drug—and to produce predictive biomarkers, prognostic biomarkers, patient-trajectory models, and other RNA-based signals for precision-oncology workflows.

The company delivers this capability through collaborations with pharmaceutical and diagnostic companies, using clinical tumor RNA-seq for model training, evaluation, and application-specific work. Its PacBio collaboration adds up to 100 fresh-frozen tumor samples sequenced with HiFi long reads, which capture transcript architecture at higher resolution; the intended benefits are better patient stratification, biomarker discovery, clinical-trial design, diagnostics, and potentially fewer patients needed to establish statistical power.

Market

Blank Bio competes in the emerging AI-for-drug-development and precision-oncology market, specifically at the intersection of RNA foundation models, transcriptomics, biomarker discovery, and clinical-trial optimization. Its competitive set includes broader genomic foundation models such as the Arc Institute's Evo 2, which performs prediction and design across DNA, RNA, and proteins, as well as RNA-focused efforts such as Deep Genomics' BigRNA and the Helix-mRNA model. These are best viewed as adjacent technical competitors or benchmarks rather than proven direct substitutes for Blank Bio's oncology-focused patient-stratification product.

The company was founded in 2025, joined Y Combinator's Summer 2025 batch, and was listed with a six-person team. Its publicly disclosed traction includes open-source model use by Sanofi and GSK, collaboration with the Arc Institute on virtual-cell models, a strategic PacBio collaboration, and an oversubscribed $7.2 million seed round in May 2026 backed by Define Ventures, Leonis Capital, Nova Threshold, Ripple Ventures, SignalFire, Y Combinator, and others. The available evidence does not disclose revenue or establish a mature commercial business, so Blank Bio is best characterized as an early-stage company with research and industry validation rather than definitively labeled pre-revenue.