Harmonized single-cell landscape, intercellular crosstalk and tumor architecture of glioblastoma

Glioblastoma, IDH wt (GB) is an aggressive brain malignancy associated with a dismal prognosis and poor quality of life. Large-scale data integration can help to uncover unexplored tumor pathobiology. To resolve the composition of the tumor milieu and recreate the cellular map of GB, we created the ‘GBmap’, a curated resource that at its core harmonizes 16 datasets gathering 110 patients and spanning over 330 thousand cells. By transfer learning, we ‘upgraded’ our atlas reference (extended GBmap) with another ten datasets, including our own profiled GB tissues resulting in a collection of 26 studies, 240 patients, and more than 1,1 million cells. We showcase the applications of our resource in several scenarios, from reference mapping to biological discoveries. The GBmap represents a framework that allows the streamlined characterization, integration, and interpretation of new data and provides a platform for exploratory analysis, hypothesis generation, and testing. Single-nuclei transcriptome (snRNA-seq) of the primary GB was generated using the 10X genomics paltform. ***Due to patient confidentiality, raw sequencing data will be submitted to EGA**