Extended Data Fig. 10: Association of neuroblastoma cell type composition with clinical parameters.
From: Single-cell transcriptomic analyses provide insights into the developmental origins of neuroblastoma
a, Hierarchical clustering of patient samples based on abundance of fetal adrenal medullary cell populations without cycling populations determined by deconvolution with BSEQ-sc (SEQC cohort). b, Composition of neuroblastoma tumors by subgroup based on deconvolution of bulk RNA-seq data with fetal adrenal cell populations without cycling populations using BSEQ-sc (SEQC cohort). c, Hierarchical clustering of patient samples based on abundance of fetal adrenal medullary cell populations determined by BSEQ-sc (TARGET cohort). d-i, Kaplan-Meier analysis of event-free (d, f, h) or overall (e, g, i) survival in neuroblastoma patients according to the relative abundance of late neuroblasts (d–g), neuroblasts (h) or cycling neuroblasts (i) determined by deconvolution of tumors with fetal adrenal cell populations. Panels f and g show deconvolution results for the TARGET cohort, panels e and i show deconvolution results for the SEQC cohort and panels d and h show deconvolution results for the SEQC cohort when excluding cycling cells from the analysis. P-values were calculated using the two-sided log-rank test.
