Single-cell genomics is deepening our understanding of the nervous system and shedding light on the heterogeneity of its cells. This focus issue of Nature Neuroscience celebrates recent methodological and analytical advances in single-cell genomics, highlights principles of study design, and aims to encourage collaboration between cellular, molecular and systems neuroscientists.

Single-cell genomics enables comprehensive, multimodal investigation of the cellular heterogeneity and architecture of the nervous system. For example, single-cell technologies can be used to study the development and evolution of the nervous system, and the multimodal data obtained can be integrated and leveraged to infer molecular mechanisms that underlie responses to diseases and adaptive phenotypes.

Achieving the true potential of single-cell genomics research requires cooperation and robust experimental design. Our focus issue on this subject is inspired by a large collaborative effort involving more than seventy experts who leverage single-cell technologies in a plethora of ways. Their efforts have resulted in three interconnected reviews1,2,3 that present an open discourse on the power and limitations of single-cell genomics, with a focus on considerations regarding experimental design.

The first review1 focuses on building molecular atlases of cell types and states in health, disease, and diverse populations. The review discusses considerations regarding experimental design depending on the aim of the study, including common pitfalls and challenges. The authors highlight the importance of integrating single-cell and single-nucleus transcriptomic data with data from other modalities. They discuss how molecular atlases can be leveraged to generate hypotheses and infer putative regulatory mechanisms that drive cell-fate specification and cell-state transitions.

The second review2 presents best practices in computational analysis of single-cell, single-nucleus or spatial sequencing datasets. Although data analysis is performed after data collection, it should guide experimental design. The accuracy of the analysis is affected by technical and biological factors (such as cell coverage or the heterogeneity of a disease) that influence the design of the computational scheme. The authors also discuss technologies that are gaining traction (such as spatial transcriptomics and single-nucleus epigenomics), consider how data from different modalities can be integrated, and discuss how the information obtained using these technologies can expand our understanding of the nervous system.

To complete the set, our third review3 tackles the increasingly pivotal matter of validation. Single-cell genomics datasets provide extensive information that often requires validation to mitigate the risk of false discoveries. Depending on the scope of the study and the knowledge gap it aims to fill, validation approaches can range from simply confirming gene expression changes with alternative methods to assessing the function of certain cell clusters or genes in a specific context, with the latter also contributing to closing the gaps between cellular, molecular and systems neuroscience.

The enormous potential of single-cell genomics results, in part, from the open availability for further exploration of the generated datasets. Appropriate data- and code-sharing practices enable transparency and reproducibility, foster trust and collaboration within the scientific community, and contribute to the democratization of research. Nature Neuroscience is a champion of data sharing, as summarized in an earlier editorial4. We also recognize that data sharing should be ethical and sustainable5,6.

Single-cell genomics has been transformative for neuroscience. It enables researchers to profile thousands or millions of cells simultaneously and unbiasedly. Single-cell and single-nucleus sequencing, spatial transcriptomics and single-nucleus epigenomics will continue to expand our understanding of the cellular composition, tissue architecture, gene regulatory landscape and, crucially, function and dysfunction of the nervous system. We look forward to seeing how integrating the unbiased querying of the cellular and molecular composition of the nervous system with the study of cellular and gene functions, systems, circuits and behaviors will contribute to unveiling its complexity.