Table 2.
Prediction accuracy of the models for the sepsis and COVID-19 patients based on the gene ensemble noise and WGCNA eigengenes explanatory variables for the discovery (Disc.) and validation (Valid.) cohorts.
Metric | Sepsis | Sepsis/CAP | Other sepsis | COVID-19 | ||||
---|---|---|---|---|---|---|---|---|
Disc | Valid | Disc | Valid | Disc | Valid | Disc | Valid | |
Gene ensemble noise | ||||||||
bACC | 0.799 | 0.701 | 0.802 | 0.798 | 0.779 | 0.761 | 0.963 | 0.95 |
Sensitivity | 0.754 | 0.6 | 0.88 | 0.867 | 0.75 | 0.8 | 0.974 | 1 |
Specificity | 0.845 | 0.801 | 0.725 | 0.73 | 0.807 | 0.722 | 0.951 | 0.9 |
WGCNA eigengenes | ||||||||
bACC | 0.805 | 0.707 | 0.8 | 0.806 | 0.82 | 0.734 | 0.963 | 0.9 |
Sensitivity | 0.799 | 0.614 | 0.8 | 0.867 | 0.798 | 0.769 | 0.974 | 0.9 |
Specificity | 0.812 | 0.8 | 0.8 | 0.746 | 0.841 | 0.7 | 0.951 | 0.9 |