Figure 1.

Sample tree clustering and clinical trait heatmap of 70 primary and 47 recurrent glioblastoma multiforme (GBM) samples. For sample tree clustering, there were no samples with height > 300, and all samples were used for the weighted gene co-expression network analysis (WGCNA). For the construction of the clinical trait heatmap, the recurrence state contained two parts classified as with or without recurrence (with recurrence is shown in red, and without recurrence is shown in white); recurrence time contained three parts classified as non-recurrence, recurrence once, and recurrence twice (non-recurrence is shown in white, recurrence once is shown in pink, and recurrence twice is shown in red). The clinical trait heatmap shows the information of all the traits. All the traits could be used for WGCNA