Review Charts How Single-Cell and Spatial Multi-Omics Methods Are Advancing Brain Research

Studying the brain through bulk tissue measurements alone leaves much of its biology out of view. The brain is made up of highly diverse cell populations arranged in complex anatomical and microenvironmental niches, a structure that conventional approaches struggle to capture. A review published in EXO – Beyond the Cell examines how single-cell and spatial multi-omics technologies are helping researchers address this gap by combining multiple molecular layers while keeping cellular or tissue context intact.

The review was led by researchers from the University of Campinas (UNICAMP), the Federal University of São Paulo, and collaborating institutions in Brazil. It surveys recent work spanning transcriptomics, epigenomics, proteomics, metabolomics, and spatial profiling.

The authors describe how single-cell multi-omics can uncover regulatory relationships that transcriptomic data alone cannot resolve. In Alzheimer's disease research, for instance, combined epigenomic and transcriptomic analyses have linked changes in chromatin organization to neuronal vulnerability and disease progression. In Parkinson's disease and psychiatric disorders, multi-omic studies have helped define cell-type- and region-specific changes involving neurons, microglia, and oligodendrocytes.

Spatial technologies add a further dimension by showing where these molecular states occur within tissue. Methods ranging from sequencing-based platforms to high-resolution imaging and mass spectrometry imaging can distinguish molecular environments across brain regions and diseased areas. In stroke, this kind of spatial mapping has revealed clear differences between the lesion core, the surrounding peri-lesional region, and tissue that remains relatively intact.

The review cautions that higher resolution doesn't automatically lead to better biological understanding. Dissociation bias, variability in postmortem tissue, data sparsity, segmentation errors, limited molecular coverage, high costs, and inconsistencies between analytical pipelines can all affect how results are interpreted and reproduced. These issues carry particular weight in brain tissue, where long neuronal processes and intricate architecture make it difficult to assign molecular signals to individual cells with confidence. 

Looking forward, the authors point to several emerging directions: artificial intelligence for integrating data across different molecular modalities, single-cell proteomics and metabolomics, morphomics, and spatiotemporal methods that can track molecular processes over time. They also call for multi-omic datasets to better represent biologically and environmentally diverse populations.

Rather than treating multi-omics as simply a set of increasingly advanced tools, the review frames its value as the ability to connect molecular state, spatial context, cellular phenotype, and disease progression. The authors note that standardized workflows, reproducibility, scalability, and independent validation will all be necessary to move these methods from research settings into neuroscience and clinical practice.