Interactive MERFISH data browser
“Seeing is Believing”
The Qian Lab uses spatial transcriptomics, a technique that maps gene activity while preserving a cell’s location, to create a high-resolution atlas of the developing human cerebral cortex. We found that its layered structure emerges much earlier than previously thought and that different brain areas form through distinct patterns—some gradually, others with sharper boundaries. These findings reshape our understanding of how the cortex is built and provide a foundation for studying neurodevelopmental disorders linked to disrupted brain organization.
A spatially resolved single-cell atlas of human cerebral cortex development
A spatial atlas of human fetal cerebral cortex development, encompassing 18 million single cells, spans eight cortical areas across four time points in the second and third trimesters. We utilized multiplexed error-robust fluorescence in situ hybridization (MERFISH), augmented with deep-learning-based cell segmentation, to examine the molecular, cellular, and cytoarchitectural development of human fetal cortex with spatially resolved single-cell resolution. We analyzed fetal human tissues from six individuals across gestational weeks (GW) 15, 20, 22, and 34, covering eight major cortical areas along the anterior-posterior (A-P) axis. We curated a panel of 300 genes for MERFISH analysis. This panel included canonical marker genes for major cell types, alongside genes selected for their cluster-specific enrichment in a published single-cell RNA sequencing (scRNAseq) dataset of the mid-gestation human fetal cortex. In total, we analyzed approximately 18 million single cells that met quality control criteria, and we integrated all experiments to cluster the cells based on their gene expression. Following a hierarchical strategy, we identified 8 cell classes, 33 cell types and 114 subtypes.
Refer to our recent publication for more detail.
Experiments included in the initial integrated analysis
| Sample | GW | Region | #Genes |
|---|---|---|---|
| UMB1117-F1b | 15 | PFC | 300 |
| UMB1117-F2a | 15 | PMC/M1 | 300 |
| UMB1117-F2b | 15 | PMC/M1 | 300 |
| UMB1117-P1 | 15 | Par | 300 |
| UMB1117-T1 | 15 | Temp | 300 |
| UMB1117-O1 | 15 | Occi | 300 |
| FB080-F1 | 20 | PFC | 300 |
| FB080-F2a | 20 | PFC | 300 |
| FB080-F2b | 20 | PFC | 300 |
| FB080-P1b | 20 | Par | 300 |
| FB080-P1a | 20 | Par | 300 |
| FB080-P2 | 20 | Par | 300 |
| FB080-T1 | 20 | Temp | 300 |
| FB080-O1a | 20 | Occi | 300 |
| FB080-O1b | 20 | Occi | 300 |
| FB080-O1c | 20 | Occi | 300 |
| FB080-O1d | 20 | Occi | 300 |
| FB121-F1 | 20 | PFC | 300 |
| FB121-F2 | 20 | PMC/M1 | 300 |
| FB121-P1 | 20 | Par | 300 |
| FB121-T1 | 20 | Temp | 300 |
| FB121-P2 | 20 | Par | 300 |
| FB121-O1 | 20 | Occi | 300 |
| FB123-F1 | 22 | PFC | 300 |
| FB123-F2 | 22 | PFC | 300 |
| FB123-F3 | 22 | PMC/M1 | 300 |
| FB123-P1 | 22 | Par | 300 |
| FB123-O1 | 22 | Occi | 300 |
| FB123-O2 | 22 | Occi | 300 |
| UMB5900-BA9 | 34 | PFC | 300 |
| UMB5900-BA4 | 34 | PMC/M1 | 300 |
| UMB5900-BA123 | 34 | Par | 300 |
| UMB5900-BA40a | 34 | Par | 300 |
| UMB5900-BA40b | 34 | Par | 300 |
| UMB5900-BA22 | 34 | Temp | 300 |
| UMB5900-BA18 | 34 | Occi | 300 |
| UMB5900-BA17 | 34 | Occi | 300 |
Additional experiments added subsequently (clustered separately)
| Sample | GW | Region | #Genes |
|---|---|---|---|
| UMB1367-P1 | 15 | Par | 300 |
| UMB1367-O1 | 15 | Occi | 300 |
| UMB1759-O1 | 18 | Occi | 960 |
| UMB1031-O1 | 20 | Occi | 960 |
| UMB1932-O1 | 21 | Occi | 960 |
| UMB5958-BA17 | adult, 22 years old | Occi | 300 |