Blank Bio Raises $7.2M Seed to Build RNA Foundation Models for Precision Oncology

Blank Bio, a San Francisco-based applied AI research lab founded in 2025, has raised $7.2M in seed funding backed by Define Ventures, Leonis Capital, Nova Threshold, Ripple Ventures, SignalFire, Y Combinator, and others. The company is building RNA foundation models designed to extract predictive biological signals from tumor transcriptomes for oncology applications including biomarker discovery, patient stratification, diagnostics, and disease trajectory modeling.

RNA foundation models are large-scale machine learning systems trained on transcriptomic data to identify biological patterns linked to disease progression and treatment response. Blank Bio believes modern oncology workflows compress too much molecular detail into simplified gene-level summaries, leaving clinically relevant biological signals behind. The financing round reflects a larger market shift happening across Healthcare AI and TechBio infrastructure, where investors are increasingly backing startups focused less on generalized healthcare automation and more on biological data infrastructure, sequencing fidelity, and computational interpretation.

What Happened

Blank Bio was founded by Jonathan “Jonny” Hsu, Philip Fradkin, and Ruian “Ian” Shi. The company currently lists a team size of 6 employees through Y Combinator materials tied to its S25 batch participation. The startup’s core thesis centers on transcriptomics, the study of RNA activity inside cells, where standard RNA-seq workflows often reduce transcriptomic information into simplified gene-level outputs because those systems are easier to operationalize across research and clinical environments.

Blank Bio is attempting to preserve richer biological detail tied to isoform architecture, mutational complexity, and patient-specific transcriptomic variation. That distinction matters because oncology remains filled with treatment variability that existing prediction systems still struggle to explain. Two patients with nearly identical clinical profiles can experience radically different treatment responses, and Blank Bio believes RNA foundation models can help uncover some of the molecular signals driving those differences.

The company also announced a strategic collaboration with Pacific Biosciences, better known as PacBio. Blank Bio will use PacBio HiFi long-read sequencing technology to generate bulk RNA sequencing data from up to 100 fresh frozen tumor samples across multiple cancer indications. Sequencing will take place at Seattle Children’s Research Institute using the SPTLabtech firefly+ platform. Long-read sequencing matters because it captures more complete transcript structures than traditional short-read sequencing workflows, potentially producing stronger predictive performance across clinical and pharmaceutical applications.

Why Blank Bio Matters

Healthcare AI spent years drowning in abstraction. A large portion of the market built software layers around healthcare operations before fully solving the biological interpretation problem underneath them. That created an ecosystem filled with administrative optimization tools wrapped in predictive language ambitious enough to make compliance officers sweat through their blazers.

Blank Bio is approaching the market from the opposite direction by focusing directly on molecular interpretation infrastructure instead of another workflow layer for healthcare systems. That positioning places Blank Bio inside a growing category of AI-native TechBio startups building foundation models for biology itself rather than simply applying generic AI tooling around healthcare operations.

RNA sits at the center of this shift because it reflects active cellular behavior in real time. DNA provides the blueprint while RNA shows what cells are actively doing with it. That distinction creates enormous implications for precision oncology, where treatment selection, biomarker development, patient stratification, and disease progression modeling all depend on understanding biological variability at increasingly granular levels.

The Competitive Landscape Around RNA Foundation Models

Blank Bio enters a market rapidly filling with AI-native biotech companies, sequencing infrastructure providers, pharmaceutical partnerships, and research institutions racing to build large-scale biological models. Companies like Recursion and Deep Genomics helped establish the broader AI biotech category, while organizations tied to Google DeepMind and NVIDIA-backed biological AI initiatives accelerated investor interest in computational biology infrastructure.

Blank Bio’s differentiation appears intentional: stay narrowly focused on RNA foundation models and precision oncology instead of attempting to become a universal biology platform overnight. That restraint may become strategically important because early-stage infrastructure companies often lose focus trying to build horizontally before establishing technical credibility inside a single domain.

Blank Bio’s narrower concentration on transcriptomic prediction and oncology workflows gives the company a clearer commercialization pathway while the broader biological foundation model market continues taking shape. Investors increasingly favor smaller technical teams capable of building specialized infrastructure faster than larger organizations buried under institutional process and fragmented decision-making.

Why Investors Are Paying Attention to TechBio Infrastructure

The investment landscape around AI biotech has changed significantly over the last 24 months. Earlier healthcare AI cycles rewarded broad automation narratives tied to administrative workflows and operational efficiency. Investors are now moving deeper into infrastructure-heavy companies with stronger technical defensibility tied directly to biological data, sequencing systems, and model development.

Blank Bio sits directly inside that transition. The PacBio collaboration also reflects another market reality emerging across genomics infrastructure where sequencing quality increasingly matters as much as model architecture. Foundation models trained on incomplete or compressed biological data risk inheriting the same analytical limitations researchers have struggled with for years.

The challenge is no longer access to biological data. The challenge is extracting clinically meaningful signal from overwhelming biological complexity. That shift is forcing both investors and operators to think less about generalized AI narratives and more about the infrastructure layers capable of supporting meaningful biological interpretation at scale.

What This Signals for Precision Oncology

Precision oncology is moving toward higher-resolution biological interpretation at the exact moment machine learning systems are becoming capable enough to process transcriptomic complexity at scale. Sequencing costs continue falling, computational infrastructure continues improving, and investors continue funding specialized AI-native biology teams capable of operating closer to the molecular layer itself. Blank Bio is positioning itself directly inside that convergence.

That does not guarantee success. Biology has a long history of humbling investors, founders, and researchers convinced they solved complexity right before the complexity introduced itself properly. Clinical validation cycles remain difficult, and translating transcriptomic prediction into real-world medical utility remains one of the hardest problems in healthcare technology.

But the direction of the market is becoming increasingly difficult to ignore. The next generation of healthcare AI companies may look less like enterprise SaaS vendors and far more like computational biology infrastructure firms capable of turning molecular chaos into clinically useful prediction.

Frequently Asked Questions

What is Blank Bio?

Blank Bio is a San Francisco-based applied AI research lab building RNA foundation models for precision oncology and clinical prediction applications.

How much funding did Blank Bio raise?

Blank Bio raised $7.2M in seed funding from Define Ventures, Leonis Capital, Nova Threshold, Ripple Ventures, SignalFire, Y Combinator, and others.

Who founded Blank Bio?

Blank Bio was founded by Jonathan “Jonny” Hsu, Philip Fradkin, and Ruian “Ian” Shi in 2025.

What are RNA foundation models?

RNA foundation models are machine learning systems trained on transcriptomic data to identify biological patterns linked to disease progression and treatment response.

Why is Blank Bio partnering with PacBio?

Blank Bio is using PacBio HiFi long-read sequencing technology to generate higher-resolution RNA sequencing data for oncology-focused foundation models.

What market is Blank Bio targeting?

Blank Bio is focused on precision oncology, biomarker discovery, patient stratification, diagnostics, and clinical trial optimization.

Why are investors funding AI-native biotech infrastructure?

Investors increasingly view biological foundation models, sequencing infrastructure, and transcriptomic analysis as foundational layers for next-generation healthcare AI systems.