biostudies-arrayexpressOliver GouldHomo sapiensSpace Ranger (Software Suite)https://www.ebi.ac.uk/biostudies/studies/E-MTAB-17164Cancer cells display heterogeneous and dynamic states in glioblastoma, but how these malignant states arise and whether they follow a tractable cellular trajectory across tumours is poorly understood. Here, we generate a deep single cell and spatial multi-region atlas of 12 isocitrate dehydrogenase wild-type (IDH-wt) primary glioblastomas that integrates transcriptomic, epigenomic and genomic analysis to comprehensively characterise their tumour heterogeneity. This submission contains the Space Ranger processed outputs from Visium spatial transcriptomic sequencing (10x Genomics), including paired high-resolution H&E tissue images. We also provide an integrated single cell and spatial dataset across 97 Visium sections, including gene expression, cell state abundances, and histopathological annotations.biostudies-arrayexpressNucleic Acid Extraction - Visium Spatial - Spatial Gene Expression for Fresh Frozen (Direct Placement): The 10x Genomics Visium Spatial Gene Expression protocol was applied to fresh frozen OCT-embedded samples. Sections were cut at 10 µm onto pre-chilled Visium Spatial Gene Expression (GE) slides (catalog no. 2000233, 10x Genomics) using a Leica CX3050S cryostat, mounting onto the 0.42 cm2 capture areas (two sections per tumour block; blocks larger than the capture area were scored and split across capture areas). Following the manufacturer's instructions (Visium Spatial Gene Expression User Guide, CG000239 Rev A), sections were fixed with cold methanol, H&E stained and imaged on a Hamamatsu NanoZoomer 2.0HT before permeabilization (most blocks processed for 12-24 mins), reverse transcription and cDNA synthesis using a template-switching protocol.Library Construction - Visium Spatial - Spatial Gene Expression for Fresh Frozen (Direct Placement): Second-strand cDNA liberated from the slide was prepared into dual-indexed libraries following library construction detailed in the Visium Spatial Gene Expression User Guide (CG000239 Rev A, 10x Genomics).Sequencing - Visium Spatial - Spatial Gene Expression for Fresh Frozen (Direct Placement): Libraries were sequenced with paired-end dual-indexing on an Illumina NovaSeq 6000, with the sequencing format read 1: 28 cycles, i7 index: 10 cycles, i5 index: 10 cycles and read 2: 90 cycles, targeting a minimum of 50,000 read pairs per spot.Nucleic Acid Extraction - Visium CytAssist Spatial Gene Expression 11mm, Human Transcriptome, Fresh Frozen: The 10x Genomics Visium CytAssist protocol was applied to fresh frozen OCT-embedded samples. 10 µm cryosections were mounted on Superfrost™ Plus Microscope Slides (Fisherbrand™). Slides were dried at RT for 5 min and fixed in cold methanol at -20°C for 30 min, then H&E stained and imaged on a Hamamatsu NanoZoomer 2.0HT. After destaining, human whole transcriptome probe pairs were hybridised and ligated to the tissue RNA. The CytAssist instrument was used to transfer the transcriptomic probes from the standard glass slide to the Visium CytAssist Spatial Gene Expression Slide (v2, 11 mm capture area), and the probes were extended to incorporate the spatial barcodes. Probe hybridization, ligation, release and extension followed the Visium CytAssist Spatial Gene Expression Reagent Kits User Guide (CG000495).Library Construction - Visium CytAssist Spatial Gene Expression 11mm, Human Transcriptome, Fresh Frozen: Released, barcoded probe products were pre-amplified and prepared into dual-indexed libraries following library construction detailed in the Visium CytAssist Spatial Gene Expression Reagent Kits User Guide (CG000495, 10x Genomics).Sequencing - Visium CytAssist Spatial Gene Expression 11mm, Human Transcriptome, Fresh Frozen: Libraries were sequenced with paired-end dual-indexing on an Illumina NovaSeq 6000, with the sequencing format read 1: 28 cycles, i7 index: 10 cycles, i5 index: 10 cycles and read 2: 90 cycles, targeting a minimum of 70,000 read pairs per spot.Sample Collection - Patients with suspected GB were identified pre-operatively and consented for entry into the study. Surgery was performed at Cambridge University Hospitals NHS Foundation Trust. Written and informed consent was obtained in accordance with the guidelines in The Declaration of Helsinki 2000. Ethical approval for the use of these tissues was obtained from the Cambridge Local Research Ethics Committee (REC 18/EE/0172). All patients underwent 5-ALA guided tumour resection as per local protocols. During tumour debulking, regions of high fluorescence were identified, their spatial location recorded and the tissue samples were collected for this study. Tissue was sampled from multiple sites of each GB tumour, targeting superior, anterior, posterior and middle regions where possible. Each tissue sample was immediately washed in saline buffer and embedded in OCT medium (Scigen OCT Compound, #4586) using a dry ice-cooled bath of isopentane at −75 °C. OCT-embedded samples were sectioned using a cryostat (Leica CX3050S). The fresh frozen tissue blocks were trimmed until the tissue surface was fully exposed. Two to three 10 µm thick sections were collected to check RNA integrity and another section was processed for Haematoxylin and Eosin (H&E) staining to assess tissue morphology. The RNA quality of each sample was evaluated by Tapestation (Tapestation RNA ScreenTape, Agilent) after isolating RNA (Qiagen EZ2 Automated RNA Extraction using EZ2 RNA/miRNA Tissue Kit Cat. No.959035). Only samples with RNA integrity number (RIN) values >7 were used for omic profiling.MIAME ScoreOrganizationAssays and DataProcessed DataMAGE-TAB FilesData Transformation - Visium spatial RNA-sequencing data, together with the high-resolution brightfield H&E images acquired on the Hamamatsu NanoZoomer 2.0HT, were used as input for the 10x Genomics Space Ranger Software Suite (version detailed in experimental variables table). Reads were mapped to the GRCh38-2020-A reference, aligning the barcoded spot pattern to the H&E tissue image and differentiating tissue from background. The resulting spot-by-gene count matrices were used for further downstream analysis, including cell type mapping with cell2location.UnknownTranscriptomicsIllumina NovaSeq 6000Cancer cells display highly heterogeneous and plastic states in glioblastoma, an incurable brain tumour. However, how these malignant states arise and whether they follow defined cellular trajectories across tumours is poorly understood. Here, we generated a deep single cell and spatial multi-omic atlas of human glioblastoma that pairs transcriptomic, epigenomic and genomic profiling of 12 tumours across multiple regions. We identify that glioblastoma heterogeneity is driven by spatially-patterned transitions of cancer cells from developmental-like states towards those defined by a glial injury response and hypoxia. This cellular trajectory regionalises tumours into distinct tissue niches and manifests in a molecularly conserved manner across tumours as well as genetically distinct tumour subclones. Moreover, using a new deep learning framework to map cancer cell states jointly with clones in situ , we show that tumour subclones are finely spatially intermixed through glioblastoma tissue niches. Finally, we show that this cancer cell trajectory is intimately linked to myeloid heterogeneity and unfolds across regionalised myeloid signalling environments. Our findings define a stereotyped trajectory of cancer cells in glioblastoma and unify glioblastoma tumour heterogeneity into a tractable cellular and tissue framework.spatial transcriptomics by high-throughput sequencingHomo sapiensA spatiotemporal cancer cell trajectory underlies glioblastoma heterogeneityOliver GouldGrant de Jong, Fani Memi, Tannia Gracia, Olga Lazareva, Oliver Gould, Georgios Solomou, Alexander Aivazidis, Manas Dave, Qianqian Zhang, Melanie Jensen, Ahmet Sureyya Rifaioglu, Joao D. Barros-Silva, Sabine Eckert, Di Zhou, Yvette Wood, Elizabeth Tuck, Sezgin Er, Koen Rademaker, Robert Petryszak, Henry Marshall, Kenny Roberts, Andrew L. Trinh, Shreya Rai, Tyler Shaw, Agnes Oszlanczi, Hayden Powell, Pasha Mazin, Stanislaw Makarchuk, Yinshui Chang, Katarzyna Kania, Liying Jin, Katy Tudor, Hon Man Chan, Harriet Johnson, Benjamin Rumney, Holly Anderson, Jasmine Halliwell, Vanessa Brito Pereira, Zoi Katsirea, Irfan Mamun, Ilaria Mulas, Annelies Quaegebeur, Mayen Briggs, Jessica Cox, Jimmy Tsz Hang Lee, Laura Rueda-Gensini, Manu Saraswat, Zaira Seferbevoka, Adam Young, Minal Patel, Tarryn Porter, Moritz Gerstung, Martin Prete, Elena Prigmore, Moritz Mall, Harry Bulstrode, Julio Saez-Rodriguez, James Briscoe, David H. Rowitch, Richard Mair, Sam Behjati, Oliver Stegle, Omer Ali BayraktarFani MemiGrant de JongTannia GraciaTong LiOmer BayraktarfalseGBM-Space: Spatial Transcriptomic Profiling of Glioblastoma (10x Genomics - Visium)Cancer cells display heterogeneous and dynamic states in glioblastoma, but how these malignant states arise and whether they follow a tractable cellular trajectory across tumours is poorly understood. Here, we generate a deep single cell and spatial multi-region atlas of 12 isocitrate dehydrogenase wild-type (IDH-wt) primary glioblastomas that integrates transcriptomic, epigenomic and genomic analysis to comprehensively characterise their tumour heterogeneity. This submission contains the Space Ranger processed outputs from Visium spatial transcriptomic sequencing (10x Genomics), including paired high-resolution H&E tissue images. We also provide an integrated single cell and spatial dataset across 97 Visium sections, including gene expression, cell state abundances, and histopathological annotations.2026-06-12T00:00:00Z2026-06-23T05:10:23.391Z2026-06-23T04:12:29.649ZE-MTAB-17164EGAD00001015527EGAS00001005801EFO_0002944EFO_0004170EFO_0030005EFO_0005518EFO_0003816EFO_000418410.1101/2025.05.13.653495