Parse Biosciences and Nomic Bio Partner to Develop Cascade, a 100 Million-Cell Dataset for AI-Driven Immune Research
Parse Biosciences, a leading provider of scalable and accessible single-cell sequencing solutions and a QIAGEN company, has partnered with Nomic Bio, a leader in high-throughput proteomic profiling, to develop Cascade, a foundational dataset comprising 100 million cells designed to deepen understanding of the human immune response. The collaboration aims to provide pharmaceutical and biotechnology companies with comprehensive cellular- and protein-level insights into immune biology, helping accelerate the identification and validation of potential drug targets.
Cascade is being developed to address a critical limitation in artificial intelligence (AI)-driven biological research: the lack of large-scale, high-quality datasets that capture how human immune cells respond to different biological signals and therapeutic interventions. By integrating single-cell transcriptomic data with quantitative proteomic measurements, the initiative is expected to provide researchers with a detailed view of immune activity, cellular changes and communication between cells over time.
The project has already attracted foundational partner commitments from leading pharmaceutical and biotechnology companies, including GSK, Pfizer, Sanofi, Graph Therapeutics and Turbine. Their participation highlights the growing interest in large-scale, multimodal biological datasets that can support AI-powered drug discovery, improve the understanding of complex biological systems and strengthen the development of predictive models.
Through this collaboration, Parse Biosciences and Nomic Bio aim to create a resource that researchers can use to investigate immune responses under a broad range of experimental conditions. The resulting dataset is expected to support the development and evaluation of AI models capable of predicting how immune cells behave when exposed to cytokines, drugs and combinations of different biological perturbations.
Cascade to Combine 100 Million Single-Cell Profiles With Proteomic Measurements
Cascade is designed as a multimodal perturbation dataset that captures how immune cells respond to treatments and biological signals across multiple timepoints. Its planned scale and breadth are intended to make it a major resource for immune system modelling and AI-enabled biological research.
The dataset will combine 100 million single-cell RNA sequencing (RNA-seq) profiles with more than 11,000 proteomic secretome profiles. These measurements will capture immune cell responses to more than 150,000 individual and combinatorial cytokine and drug perturbation conditions.
The combination of these data types is central to Cascade’s research objectives. Single-cell RNA sequencing enables scientists to examine gene expression in individual cells, revealing differences in cellular states, molecular pathways and responses to external stimuli. This level of resolution can help researchers understand why cells of the same general type may react differently to the same treatment.
Proteomic secretome profiling, meanwhile, measures proteins secreted by cells into their surrounding environment. These proteins can provide important information about cellular communication, immune activation, inflammatory processes and other biological mechanisms involved in immune responses.
By bringing transcriptomic and proteomic measurements together, Cascade is intended to provide a more complete picture of immune biology than either data type could offer independently. Researchers will be able to examine changes in gene expression alongside the proteins associated with cellular signalling and communication.
The inclusion of multiple timepoints will further support investigations into how immune responses develop. Rather than capturing only a single snapshot of cellular activity, the dataset is designed to reveal how responses change over time following exposure to specific perturbations.
This time-resolved approach could help researchers distinguish early cellular responses from later changes in gene expression and protein secretion. It may also provide insights into how different immune cell populations interact and how their responses evolve under individual or combined treatments.
The dataset’s extensive range of experimental conditions is expected to offer researchers opportunities to investigate both established and less well-understood relationships within the immune system. Such information could help scientists formulate hypotheses about biological pathways, identify promising therapeutic targets and evaluate how specific interventions influence immune activity.
Addressing the Data Bottleneck in Biological AI
AI is increasingly being applied to biological research to identify patterns across complex datasets, generate predictions about cellular behaviour and support the discovery of new medicines. However, the performance of these models depends heavily on the availability, quality and diversity of the data used to train and evaluate them.
Immune biology presents a particularly challenging area for AI modelling because immune responses involve numerous cell types, molecular signals and interacting pathways. A single intervention can trigger multiple downstream effects, while different cell populations may respond in distinct ways depending on their biological state and the surrounding environment.
Despite the importance of these processes in drug discovery, comprehensive datasets measuring human immune responses to key perturbations remain limited. This creates a challenge for researchers seeking to develop predictive models that can reliably estimate how cells will respond to new drugs or combinations of treatments.
Cascade is intended to help address this limitation by generating a large, structured dataset focused on immune responses. Its combination of single-cell gene expression, secreted protein measurements and multiple perturbation conditions could give AI developers a broader evidence base for learning relationships between interventions and biological outcomes.
Charlie Roco, Co-founder and Chief Technology Officer at Parse Biosciences, highlighted the importance of addressing the data limitations affecting biological AI.
“Biological AI modeling has hit a data bottleneck. Despite their central role in drug discovery, there is a surprisingly small amount of available data measuring the human’s immune response to key perturbations like cytokines and other drugs. Cascade is a critical step to revealing these interactions, and we expect to dramatically improve the predictive power of our partner’s AI models as a result,” Roco said.
His comments reflect the premise behind the project: that larger and more informative datasets could improve the ability of AI systems to identify meaningful biological relationships and generate more useful predictions.
By providing measurements across a wide range of perturbations, Cascade could support the development of models that predict cellular responses to conditions not directly represented in the training data. Such capabilities would be valuable for prioritising experiments, exploring potential drug targets and narrowing the range of therapeutic candidates that require further laboratory investigation.
However, the utility of these models will depend on the quality of the experimental data, the robustness of model development and independent validation of predictions. Cascade is being designed to provide a substantial data resource for these efforts, rather than replacing the experimental work needed to establish biological and therapeutic relevance.
Nomic Bio to Contribute High-Throughput Proteomic Profiling
Nomic Bio will contribute proteomic measurements through its Omini 1000 platform, which is designed to measure more than 1,000 secreted proteins across all samples.
The platform’s role is important because gene expression alone does not fully capture the complexity of immune signalling. Although RNA sequencing can reveal which genes are active within individual cells, protein measurements can provide complementary information about the molecules released into the extracellular environment.
Secreted proteins play important roles in communication between cells and can influence a wide range of immune processes. Measuring these proteins alongside gene expression may help researchers better understand how cells respond to perturbations and how their activity affects other cells.
The integration of quantitative proteomic data with single-cell transcriptomic profiles is intended to improve the ability of AI models to represent these interactions. It may also help researchers investigate relationships between changes in cellular gene expression and the signals associated with immune activation, regulation and communication.
Milad Dagher, Co-founder and Chief Executive Officer of Nomic Bio, emphasised the importance of incorporating protein-level information into models of immune biology.
“AI models of immune biology need to capture not only how cell states change, but also the protein signals cells use to communicate,” Dagher said. “Cascade brings quantitative, time-resolved protein measurements together with single-cell data at a scale that can help researchers model how immune responses develop and predict how they change under new perturbations.”
Through its contribution, Nomic Bio will help extend the dataset beyond measurements of gene activity to include a broader range of molecular signals associated with immune responses. This integrated approach is intended to provide a stronger foundation for understanding the relationships between cellular states and secreted proteins.
Parse Biosciences to Generate Single-Cell Data Through GigaLab
Parse Biosciences will oversee the generation of the single-cell RNA sequencing data through its GigaLab, a high-throughput service laboratory capable of profiling billions of cells annually.
The facility’s capacity is intended to support the substantial scale of the Cascade project. Generating 100 million single-cell profiles requires infrastructure and workflows capable of processing large numbers of samples while maintaining consistent data quality.
High-throughput single-cell sequencing allows researchers to examine cellular heterogeneity at a scale that would be difficult to achieve using conventional approaches. Instead of averaging gene expression across an entire population, scientists can analyse individual cells and identify distinct populations, response patterns and molecular characteristics.
For immune research, this capability is particularly relevant because immune cell populations can contain diverse subgroups with different functions and activation states. Detailed profiling can help researchers investigate how these subgroups respond to different treatments and how those responses vary across experimental conditions.
By combining the scale of Parse’s GigaLab with Nomic Bio’s proteomic profiling platform, the partners aim to create a dataset that brings together complementary forms of biological information within a common research resource.
The collaboration is also intended to make the resulting data available to pharmaceutical and biotechnology partners seeking to strengthen their AI-driven research programmes. Access to large-scale, multimodal data could help these organisations investigate immune mechanisms, evaluate potential drug targets and improve the evidence supporting early-stage drug discovery decisions.
Pharmaceutical and Biotechnology Partnerships Highlight Industry Interest
The initial commitments from GSK, Pfizer, Sanofi, Graph Therapeutics and Turbine demonstrate interest in the development of large-scale datasets for immune modelling and drug discovery.
Pharmaceutical and biotechnology companies increasingly need tools that can help them understand complex biological systems and identify promising therapeutic opportunities. Data-driven approaches may help researchers examine more hypotheses, prioritise experiments and improve the efficiency of early drug development.
A resource such as Cascade could be useful across multiple stages of target discovery and validation. Researchers may use the data to examine how particular cytokines or drugs influence immune cell populations, investigate potential molecular mechanisms and develop computational models for predicting responses to new interventions.
The dataset could also help AI developers compare predicted biological outcomes with experimentally measured results, supporting efforts to assess the reliability of their models.
Parse Biosciences and Nomic Bio are continuing to welcome additional partners interested in participating in the initiative. Organisations seeking licensing information can visit www.parsebiosciences.com/cascade.
Advancing Multimodal Data Resources for Drug Discovery
The development of Cascade reflects a broader effort to expand the data infrastructure supporting AI-enabled life sciences research. As researchers work to model increasingly complex biological systems, the availability of large datasets that capture multiple molecular layers is becoming an important consideration.
By combining 100 million single-cell RNA-seq profiles with more than 11,000 proteomic secretome profiles across more than 150,000 individual and combinatorial perturbation conditions, Parse Biosciences and Nomic Bio aim to provide researchers with a resource for investigating immune biology at substantial scale.
The integration of transcriptomic and proteomic information could help address gaps in current understanding of immune signalling and provide a stronger foundation for predictive cellular models. These models may ultimately support more informed decisions during drug target discovery and validation, although their value will need to be established through rigorous testing and continued experimental research.
For Parse Biosciences and Nomic Bio, the collaboration brings together expertise in single-cell sequencing and high-throughput proteomic profiling to address a shared challenge in biological AI: the need for comprehensive experimental data that captures how cells respond to interventions.
As additional partners join the initiative, Cascade is intended to support further development of AI models capable of representing immune responses and exploring how those responses change under different perturbations. The project underscores the growing role of integrated biological datasets in efforts to improve the understanding of human biology and accelerate the discovery of potential new therapies.
About Nomic Bio
Nomic was founded to make measuring biology easier and enable scientists to extend lives by making proteomics accessible, scalable, and routine. Powered by its proprietary nELISA® technology, Nomic delivers large-scale, quantitative, and cost-effective protein data to accelerate discovery, development, and translation. Its end-to-end service combines flexible content, expert support, and analysis-ready outputs, enabling seamless integration of proteomics across all stages of drug and biomarker development. Nomic’s mission is to make proteomics ubiquitous for modern biology. To learn more, visit nomic.bio or follow us on LinkedIn.
About Parse Biosciences
Parse Biosciences, a QIAGEN company, is a global life sciences company whose mission is to accelerate progress in human health and scientific research. Empowering researchers to perform single cell sequencing with unprecedented scale and ease, its pioneering approach has enabled groundbreaking discoveries in cancer treatment, tissue repair, stem cell therapy, kidney and liver disease, brain development, and the immune system.