Table 2.
Recursive partitioning for DNA concentration (ng/µl) and cancer cell proportion (%) needed to reach a (a) methylation class classifier score (CS) bigger than 0.84 in the Heidelberg brain tumor classifier, (b) methylation class classifier score (CS) bigger than 0.9 in the Heidelberg brain tumor classifier, (c) family classifier score (CS) bigger than 0.9 in the Bethesda brain tumor classifier, (d) class classifier score (CS) bigger than 0.9 in the Bethesda brain tumor classifier
| DNA concentration (ng/µl) > / = 1.21 | DNA concentration (ng/µl) < 1.21 | |
|---|---|---|
| (a) Heidelberg brain tumor classifier, methylation class, calibrated score cut-off 0.84 | ||
| n | 192 | 14 |
| G^2 | 230 | 16.75 |
| CS < 0.84 probability | 0.29 | 0.69 |
| CS > 0.84 probability | 0.71 | 0.31 |
| R2 | 0.039 | |
| Split p value | 0.1037 | |
| Cancer cells (%) > / = 53 | Cancer cells (%) < 53 | |
|---|---|---|
| n | 157 | 49 |
| G^2 | 171.46 | 66.92 |
| CS < 0.84 probability | 0.24 | 0.57 |
| CS > 0.84 probability | 0.76 | 0.43 |
| R2 | 0.072 | |
| Split p value | 0.000999 |
| DNA concentration (ng/µl) > / = 3.18 | DNA concentration (ng/µl) < 3.18 | |
|---|---|---|
| (b) Heidelberg brain tumor classifier, methylation class, calibrated score cut-off 0.9 | ||
| n | 163 | 43 |
| G^2 | 205.61 | 59.03 |
| CS < 0.9 probability | 0.33 | 0.55 |
| CS > 0.9 probability | 0.67 | 0.45 |
| R2 | 0.028 | |
| Split p value | 0.3152 | |
| Cancer cells (%) > / = 62 | Cancer cells (%) < 62 | |
|---|---|---|
| n | 94 | 112 |
| G^2 | 97.31 | 155.23 |
| CS < 0.9 probability | 0.21 | 0.51 |
| CS > 0.9 probability | 0.79 | 0.49 |
| R2 | 0.073 | |
| Split p value | 0.000398 |
| DNA concentration (ng/µl) > / = 2.95 | DNA concentration (ng/µl) < 2.95 | |
|---|---|---|
| (c) Bethesda brain tumor classifier, family | ||
| n | 178 | 48 |
| G^2 | 133.15 | 59.62 |
| CS < 0.9 probability | 0.12 | 0.31 |
| CS > 0.9 probability | 0.88 | 0.69 |
| R2 | 0.043 | |
| Split p value | 0.2409 | |
| Cancer cells (%) > / = 37.4 | Cancer cells (%) < 37.4 | |
|---|---|---|
| n | 221 | 5 |
| G^2 | 189.76 | 6.73 |
| CS < 0.9 probability | 0.15 | 0.53 |
| CS > 0.9 probability | 0.85 | 0.47 |
| R2 | 0.024 | |
| Split p value | 0.8743 |
| DNA concentration (ng/µl) > / = 1.59 | DNA concentration (ng/µl) < 1.59 | |
|---|---|---|
| (d) Bethesda brain tumor classifier, class | ||
| n | 203 | 23 |
| G^2 | 201.49 | 31.49 |
| CS < 0.9 probability | 0.2 | 0.43 |
| CS > 0.9 probability | 0.8 | 0.57 |
| R2 | 0.025 | |
| Split p value | 0.6385 | |
| Cancer cells (%) > / = 79.5 | Cancer cells (%) < 79.5 | |
|---|---|---|
| n | 7 | 219 |
| G^2 | 0 | 235.31 |
| CS < 0.9 probability | 0.028 | 0.23 |
| CS > 0.9 probability | 0.97 | 0.77 |
| R2 | 0.013 | |
| Split p value | 0.975 |