Table 1.
Terms and definitions used in the analysis between in silico predictions and in vivo observations
| Term | Definition | Description |
|---|---|---|
| True postive (TP) |
In silico: viable phenotype |
Results where the metabolic model predicts a viable phenotype and the actual phenotype is also viable |
| |
In vivo: viable phenotype |
|
| True negative (TN) |
In silico: lethal phenotype In vivo: lethal phenotype |
Results where the metabolic model predicts a lethal phenotype and the actual phenotype is also lethal |
| False positive (FP) |
In silico: viable phenotypeIn vivo: lethal phenotype |
Results where the metabolic model predicts a viable phenotype but the actual phenotype is lethal |
| False negative (FN) |
In silico: lethal phenotype In vivo: viable phenotype |
Results where the metabolic model predicts a lethal phenotype but the actual phenotype is viable |
| Overall accuracy |
(TP + TN)/(TP + TN + FP + FN) |
Percentage of correct predictions by the metabolic model |
| Negative prediction rate |
(TN)/(TN + FN) |
Percentage of negative predictions that are correctly predicted as lethal |
| Positive prediction rate |
(TP)/(TP + FP) |
Percentage of positive predictions that are correctly predicted as viable |
| Sensitivity |
(TP)/(TP + FN) |
Percentage of viable predictions correctly predicted as positive |
| Specificity | (TN)/(TN + FP) | Percentage of lethal predictions correctly predicted as negative |