Skip to main content
Fig. 3 | BMC Cancer

Fig. 3

From: Sequential analysis of transcript expression patterns improves survival prediction in multiple cancers

Fig. 3

Sequential t-SNE analysis of OV and UCEC. a. t-SNE-generated patterns of Pyrimidine Biosynthesis Pathway transcripts showing four OV tumor clusters. b. Kaplan-Meier survival curves of patients from each of the Clusters depicted in A. c. t-SNE-generated patterns of OV Cell Cycle Pathway transcripts. d. Kaplan-Meier survival curves of patients from each of the Clusters shown in C. e. Kaplan-Meier survival curves of patients from unfavorable survival Cell Cycle Pathway Cluster 3 from C after t-SNE re-analysis using transcripts from the Pyrimidine Biosynthesis Pathway. Four t-SNE clusters similar to those depicted in A were generated (not shown) with two of these (Clusters 2 and 4) showing significant survival differences. f. Favorable survival Cell Cycle Pathway t-SNE Cluster 2 from (D) was analyzed with Pyrimidine Biosynthesis Pathway transcripts. Of these, Cluster 1 (median survival 1736 days) had a more favorable long-term survival than either Cluster 3 (1336 days) or Cluster 4 (1213 days) (P = 0.017 and P = 0.004, respectively. g. t-SNE-generated patterns of UCEC Myc Pathway transcripts. h. Kaplan-Meier survival of patients from the groups depicted in G. (I). t-SNE-generated patterns of UCEC Wnt Pathway transcripts. j. Kaplan-Meier survival curves of patients from each of the groups depicted in I. k. Kaplan-Meier survival curves of individuals from favorable survival Wnt Pathway t-SNE Cluster 2 following subsequent t-SNE profiling with Myc Pathway transcripts. l. Kaplan-Meier survival curves of patients from the unfavorable survival Wnt Pathway Cluster 1 cohort in K following repeat t-SNE profiling with Myc Pathway transcripts

Back to article page