Extended Data Fig. 5: Characterization and interpretation of GSC transcriptional gradient. | Nature Cancer

Extended Data Fig. 5: Characterization and interpretation of GSC transcriptional gradient.

From: Gradient of Developmental and Injury Response transcriptional states defines functional vulnerabilities underpinning glioblastoma heterogeneity

Extended Data Fig. 5

a, PCA plot of 69,393 cells from 29 GSC cultures. Plot colored by cell density (left). PCA plot with cells belonging to outlier sample G800_L, colored red. Remainder of cells colored grey (middle). Quantification of deviation from the mean of PC2 (y-axis) across samples. G800_L (red) represents an outlier with >95% of cells within the sample greater than two standard deviations from the mean. Horizontal dashed red line represents threshold of two standard deviations to determine outliers (right). Box plots represent the median, upper and lower quartiles of the distribution and whiskers represent 1.5-times interquartile range or the most extreme value. Outliers represented as circles. b, Correlation of cell type gene signature scores from PC1 cell embeddings (n = 65,655 cells from 28 GSC cultures; outlier G800_L removed as in Fig. 2a). Only correlations with Spearman correlation coefficient greater than |0.5| shown. Bars colored by gene signature source. c, Enriched MSigDB gene sets (FDR < 0.01) for top 100 and bottom 100 genes for PC1. (n = 65,655 cells from 28 GSC cultures; outlier G800_L removed as in Fig. 2a). d, Gene Set Enrichment Analysis (GSEA) on PC1 loadings (gene associations with PC1) visualized using EnrichmentMap (n = 65,655 cells from 28 GSC cultures; outlier G800_L removed as in Fig. 2a). Similar pathways (circles) are grouped into labeled clusters (larger bubbles). Blue circles denote positively associated pathways (Injury Response associated) and red circles denote negatively associated pathways (Developmental associated). Edges (lines) denote overlap between pathways.

Source data

Back to article page