Fig. 5: Chromatin accessibility features associated with B-ALL prognosis.

a Gene set enrichment analysis shows that 3976 target genes regulated by 70,573 RFS-related ACRs are enriched for hematopoietic development and cell cycle-associated biological process terms. The categories with the most significant FDR values are listed and are sorted by FDR values in reverse order. b T-SNE plot showing the unsupervised clustering of 42 relapse B-ALL samples based on the top 10% most significant RFS-related ACRs. B-ALL samples were clustered into Group A and Group B. c The relapse-free survival rate (RFS, n = 42) and overall survival rate (OS, n = 26) estimates for B-ALL samples of Group A and Group B. (Log-rank test, p = 3.778e−10 for RFS and p = 5.585e−7 for OS, the error bands indicate 95% confidence intervals, and the dash line indicate the median survival time). d Unsupervised clustering based on the CPM of differential ACRs between three hyperdiploidy cases in Group B and nine hyperdiploidy cases in Group A. e Gene set enrichment analysis indicates that 603 target genes regulated by 3156 ACRs up-regulated in the hyperdiploidy cases in Group B were enriched for stem cell- and myeloid progenitor-associated gene expression signatures (left panel) and cell adhesion/migration-associated biological process terms (right panel). The top categories with the most significant FDR values are listed. f Unsupervised clustering analysis of 43 hyperdiploidy cases from the TARGET project based on the expression of 603 target genes regulated by 3156 ACRs up-regulated in Group B. g Event-free survival rate (EFS, n = 35) and overall survival rate (OS, n = 36) estimates for hyperdiploidy B-ALL samples from the TARGET project. Cases in Cluster3 with high expression of the target genes and mimicking the hyperdiploidy B-ALL cases in Group B had worse outcomes. (Log-rank test, EFS p = 0.0012 and OS p = 0.0002, the error bands indicate 95% confidence intervals, and the dash line indicate the median survival time).