Parse, Nomic Bio Build 100 Million-Cell Dataset

The Cascade dataset pairs single-cell sequencing with proteomic profiling to address a data gap in biological AI modeling.

Updated on Nov. 12, 2026 in Biotech

Bold flat-color vector editorial illustration of a grid of amber-filled glass vials on a steel tray, suggesting a large standardized biological dataset.
Parse Biosciences and Nomic Bio are building Cascade, a 100-million-cell dataset backed by GSK, Pfizer, Sanofi, Graph Therapeutics and Turbine. AI Illustration. Upload story photo >

Parse Biosciences, a QIAGEN company, and Nomic Bio have partnered to develop Cascade, a foundational dataset of 100 million cells. GSK, Pfizer, Sanofi, Graph Therapeutics and Turbine have made foundational partner commitments to the project.

Why it matters

Biological AI models are constrained by limited data on immune responses to cytokines and drugs, and Cascade is designed to address that bottleneck. Parse expects the dataset to substantially improve the predictive power of partner AI models.

Cascade pairs 100 million single-cell RNA-seq profiles with more than 11,000 proteomic secretome profiles, measured on Nomic's Omini 1000 platform, which quantifies more than 1,000 secreted proteins. The dataset captures immune cell responses across more than 150,000 cytokine and drug perturbation conditions, though its delivery timeline has not been stated.

The players

Parse Biosciences

A QIAGEN company that provides scalable single-cell sequencing solutions to research and drug-development customers.

Nomic Bio

A specialist in high-throughput proteomic profiling that develops the Omini 1000 platform.

GSK

A global pharmaceutical company that made a foundational partner commitment to the Cascade dataset.

Pfizer

A global pharmaceutical company that made a foundational partner commitment to the Cascade dataset.

Sanofi

A global pharmaceutical company that made a foundational partner commitment to the Cascade dataset.

The details

Single-cell RNA-seq data will be generated at Parse's GigaLab, a high-throughput service laboratory, and paired with proteomic secretome measurements produced by Nomic's Omini 1000 platform. The combined dataset is intended to give AI developers a large, standardized resource on immune responses to cytokines and drugs.

Roadmap

Cascade extends single-cell perturbation profiling from one-off experiments into a shared, industrial-scale training resource for biological AI. It follows the broader pattern of pharmaceutical companies pooling data resources to address the data bottleneck in drug-response modeling.

For researchers and developers, Cascade could mean access to a large standardized dataset for training models of immune response to drugs. The source does not describe how the resource will be licensed or made available beyond the foundational partners.

The takeaway

Large, standardized biological datasets may matter as much as model architecture for progress in AI-driven drug discovery. Researchers tracking this space should watch how access to resources like Cascade is structured.

Further reading

Explore more biotech news on how large datasets are reshaping drug development.

More information

Learn more on the Nomic Bio company website.

Source note: This article includes information reported by Themachinemaker.