Table 3 Validation of RSKC through random baseline clustering. This table shows the scores corresponding to the best clustering result, the average and the mean value from the 1000 random datasets, comparing it with the reference scores
From: Disclosing transcriptomics network-based signatures of glioma heterogeneity using sparse methods
Case | K | Reference score | Random score (Best – Average – Median) | ||
|---|---|---|---|---|---|
Silhouette | Calinski-Harabasz | Silhouette | Calinski-Harabasz | ||
(A) LGG vs GBM | 2 | 0.70 | 640.95 | 0.72 — 0.52 — 0.52 | 668.21 — 170.93 — 158.06 |
3 | 0.62 | 746.37 | 0.59 — 0.33 — 0.34 | 247.07 — 66.21 — 60.47 | |
(B) Astrocytoma vs Oligodendroglioma | 2 | 0.58 | 513.29 | 0.74 — 0.45 — 0.46 | 264.50 — 57.91 — 46.17 |
3 | 0.59 | 526.92 | 0.60 — 0.32 — 0.31 | 61.16 — 8.24 — 6.92 | |
(C) Astrocytoma vs Oligodendroglioma vs GBM | 2 | 0.73 | 843.31 | 0.73 — 0.51 — 0.52 | 497.87 — 153.68 — 142.14 |
3 | 0.61 | 753.07 | 0.64 — 0.33 — 0.33 | 300.62 — 62.19 — 55.98 | |