Table 1 Model performance of cancer detection
From: Integrative modeling of tumor genomes and epigenomes for enhanced cancer diagnosis by cell-free DNA
Models | MGI platform | Illumina platform | ||||||
|---|---|---|---|---|---|---|---|---|
Training Cohort | Vadliation Cohort | Training Cohort | Vadliation Cohort | |||||
auc | 95% CI | auc | 95% CI | auc | 95% CI | auc | 95% CI | |
Combined | 0.984 | 0.978–0.988 | 0.902 | 0.85–0.944 | 0.946 | 0.933–0.958 | 0.926 | 0.882–0.962 |
Genome | 0.967 | 0.956–0.976 | 0.99 | 0.978–0.998 | 0.931 | 0.916–0.944 | 0.865 | 0.807–0.915 |
Epigenome | 0.938 | 0.926–0.949 | 0.842 | 0.772–0.899 | 0.877 | 0.857–0.896 | 0.918 | 0.871–0.958 |
CNV | 0.921 | 0.908–0.933 | 0.692 | 0.608–0.778 | 0.856 | 0.833–0.877 | 0.898 | 0.847–0.942 |
Fragpattern | 0.904 | 0.889–0.917 | 0.756 | 0.676–0.827 | 0.801 | 0.776–0.826 | 0.898 | 0.839–0.946 |
Fragsize | 0.835 | 0.813–0.86 | 0.852 | 0.788–0.91 | 0.866 | 0.846–0.887 | 0.904 | 0.848–0.953 |
