Fig. 10: Association of DEPTH scores with clinical features and immune signatures in validation datasets.
From: An algorithm to quantify intratumor heterogeneity based on alterations of gene expression profiles
a DEPTH scores have significant positive correlations with stemness scores, the expression levels of proliferation marker genes (MKI67, TOP2A, and RACGAP1), and proliferation signature scores in a wealth of validation datasets. b DEPTH scores have significant inverse correlations with at least three of the five immune signatures in 22 validation datasets and have significant inverse correlations with the ratios of CD8+/CD4+ regulatory T cells in 17 validation datasets (Spearman’s correlation test, p < 0.05). c Kaplan−Meier survival curves displaying the negative correlation between DEPTH scores and overall survival in multiple large-scale validation datasets. The log-rank test p-values, HR, and 95% CI are shown. d The high-grade (G3-4) tumors have significantly higher DEPTH scores than the low-grade (G1-2) tumors in multiple validation datasets (one-sided Mann–Whitney U test, p < 0.001). e The AML patients with cytogenetic abnormalities (t(9;11)(p22;q23), t(10;11)(p11.2;q23), del5q, trisomy 21, or MLL1 translocation) have significantly higher DEPTH scores than those without such abnormalities. f DEPTH scores show a positive correlation with the bone marrow leukemic blast percentages of AML patients. DEPTH scores are higher in the AML patients at the undifferentiated stage (M0) versus at the differentiated stage (M1-7) and are higher in the AML patients with other sites of relapse versus without other sites of relapse. g The high-risk cancer patients have higher DEPTH scores than low-risk cancer patients in AML and NBL. NBL, neuroblastoma. h Kaplan−Meier survival curves displaying the negative correlation between DEPTH scores and overall survival in pan-cancer. The results in (e–h) were obtained by analyzing the TARGET dataset.
