Interview Understanding the Human Brain, One Cell at a Time
In a novel international study, researchers have created the Human Neural Organoid Cell Atlas (HNOCA), a comprehensive resource mapping the cellular diversity of human neural organoids. This atlas integrates 1.7 million single-cell transcriptomic profiles, offering a powerful tool for studying brain development, disease mechanisms, and organoid fidelity. In this interview, Prof. Fabian Theis, Co-last author of the study, talks about the significance of this new resource, its potential impact on neuroscience and medicine, and the crucial role of AI in building this resource.
In a novel international study, researchers have created the Human Neural Organoid Cell Atlas (HNOCA), a comprehensive resource mapping the cellular diversity of human neural organoids. This atlas integrates 1.7 million single-cell transcriptomic profiles, offering a powerful tool for studying brain development, disease mechanisms, and organoid fidelity. In this interview, Prof. Fabian Theis talks about the significance of this new resource, its potential impact on neuroscience and medicine, and the crucial role of AI in building this resource.
The new study introduces the Human Neural Organoid Cell Atlas (HNOCA). What makes this resource a game-changer for studying the human brain and its development?
FT: Neural organoids are a highly promising technology for understanding brain development, disease and testing drugs. However, there are still many open questions in the field, for example, regarding the ability of these organoids to recapitulate different regions of the brain.
To tackle some of these pressing challenges, we teamed up with researchers from the labs of Barbara Treutlein and Gray Camp to build the HNOCA. It represents a big step towards a better understanding of these models by providing a comprehensive map of the cellular diversity in human neural organoids.
Leveraging recent advances in AI, we integrated data from over 1.7 million single-cell transcriptomic profiles across 36 datasets. This resource allowed us to systematically assess the fidelity and variability of organoids compared to the developing human brain. Specifically, the HNOCA enables the comparison of organoid protocols with respect to represented brain regions and transcriptomic fidelity.
This integrated resource addresses key challenges in the field, such as organoid-to-brain comparison and protocol standardization.
TUM | © Astrid Eckert"The Human Neural Organoid Cell Atlas represents a big step towards a better understanding of these models by providing a comprehensive map of the cellular diversity in human neural organoids."
Prof. Fabian Theis
Neural Organoids
Neural organoids are tiny, lab-grown structures designed to mimic key features of the human brain. They are created from stem cells, which can develop into different types of cells in the body. Scientists use neural organoids to study brain development, investigate neurological diseases, and test potential treatments in a controlled, ethical setting.
Neural organoids are often seen as a bridge between basic research and clinical applications. How do you envision the HNOCA contributing to advancements in understanding brain diseases or developing therapies?
FT: Neural organoids have already demonstrated exciting potential as models for brain diseases, and the HNOCA amplifies this potential by providing a reference atlas to compare disease-specific organoids against a robust control dataset. For example, it facilitates the precise annotation of cell types and regional identities, enabling researchers to identify disease phenotypes at the cellular level. By highlighting transcriptomic differences between diseased and a comprehensive collection of healthy states, the HNOCA helps uncover pathways and genes implicated in the diseases of interest.
Additionally, the atlas can guide therapeutic development by identifying further molecular targets for intervention and optimizing organoid protocols for drug screening, ensuring their relevance to human biology.
Human Cell Atlas
The Human Neural Organoid Cell Atlas is part of the international collaborative efforts to create the comprehensive Human Cell Atlas. Launched in 2016, this large international collaborative research aims to map all cell types in the healthy body, across time from development to adulthood, and eventually to old age. Creating this comprehensive reference map of human cells will guide our understanding of health and disease and drive major advances in healthcare and medicine worldwide.
Your work involves integrating vast amounts of data across labs and protocols. What role does artificial intelligence play in addressing such complex challenges, and where do you see the field heading in the next decade?
FT: Advances in artificial intelligence models have played an integral part in the creation and utilization of the HNOCA. Its development involved advanced algorithms to harmonize diverse datasets and ensure accurate annotation of cell types and states. One of the HNOCA’s most transformative applications lies in enabling generative AI models to learn and predict the effects of perturbations in organoid development. By integrating the HNOCA as a reference, AI can help to identify relevant organoid protocol optimizations, guiding the generation of cell types and brain regions that are currently underrepresented, such as certain mid-/hindbrain components or advanced maturation stages.
Looking forward, we foresee a future where virtual cell models—powered by AI trained on large-scale reference datasets like the HNOCA—become central to neuroscience research. These models could simulate the behavior of whole cell populations under various experimental conditions, dramatically accelerating the discovery of optimized differentiation protocols. Furthermore, as foundation models in biology mature, they could integrate multi-omics data, imaging, and transcriptomics to create comprehensive virtual brain organoid systems, enabling unparalleled insights into human brain development and disease. This synergy between AI innovation and biological experimentation will redefine the boundaries of biomedical research.
About the Study and the Researchers
The study was a joint effort with researchers from ETH Zurich, the Institute of Human Biology at Roche and Helmholtz Munich within the larger Human Cell Atlas (HCA) initiative. The work has now been published as part of the HCA publication bundle on the cover of Nature magazine.
Prof. Fabian Theis is Director of the Computational Health Center and the Institute of Computational Biology at Helmholtz Munich. This work was co-led by postdoctoral researcher Leander Dony and colleagues from ETH Zurich and the Insitute of Human Biology at Roche with support from doctoral researchers Artur Szałata and Irena Slišković as well as visiting Fulbright Scholar Katelyn X. Li.
Original Publication
He, Z., Dony, L., Fleck, J.S. et al. An integrated transcriptomic cell atlas of human neural organoids. Nature 635, 690–698 (2024). https://doi.org/10.1038/s41586-024-08172-8
Latest update: November 2024.
