Extended Data Fig. 2: Longitudinal cell-type changes in the TME, global similarity of cell states between primary and recurrent samples and prognostically significant cell-type profiles. | Nature Genetics

Extended Data Fig. 2: Longitudinal cell-type changes in the TME, global similarity of cell states between primary and recurrent samples and prognostically significant cell-type profiles.

From: Deciphering the longitudinal trajectories of glioblastoma ecosystems by integrative single-cell genomics

Extended Data Fig. 2

a, Relative granular cell-type abundance across the two time points (primary and recurrence). Unadjusted two-sided paired Wilcoxon rank-sum test is shown (n = 56 patients). Dotted lines connect patient values. b, Relative granular non-malignant cell state abundance across the two time points (primary and recurrence) that showed statistical significance between time points following adjustment for multiple comparisons. Unadjusted two-sided paired Wilcoxon rank-sum test is shown (n = 56 patients). Dotted lines connect patient values. c, Average proportion of cellular composition (left) of malignant cell state composition (middle) and average baseline profile score (right) for each feature group. d, Available clinical data information for this study and ref. 14 dataset. e, Comparison of the number of detected genes per cell in cells passing quality filters between this study and ref. 14 dataset. Two-sided Wilcoxon rank-sum test is shown. Box plots span from the first to third quartiles, median values are indicated by a horizontal line, whiskers show 1.5× interquartile range and outlier points are not shown. Continuous distributions of all values are shown to the right of each box plot. f, Proportion (%) of samples from ref. 14 assigned to each compositional group across the two time points (primary and recurrence). g, Kaplan–Meier curves depicting the survival time after 2nd resection according to the transcriptomic features at 2nd resection, which shows the statistical significance in the distribution between groups. Top panel is for composition group. Bottom panel is for malignant state groups. Statistical significance of survival difference between the groups in each panel was computed using the log-rank test. h, Forest plot reflects Cox proportional hazard model for time from 2nd resection to outcome, including relevant biological and clinical covariates in addition to composition group (top) and malignant state group (bottom). Each variable is presented along with its categorical possibilities and sample size on the left, while the boxes represent the hazard ratio and lines represent the 95 % confidence interval.

Back to article page