Table 1.
Detailed information regarding the 19 bacterial species investigated.
Figure 1.
Training workflow based on 18 groups of essential genes and using Ecol as test.
Figure 2.
ROC curve of cross-organism validating for Ecol.
Black diamond denotes the default cutoff of 0.15.
Figure 3.
The precision, sensitivity and specificity in relation to the cutoff rank.
The vertical dashed line represents the default cutoff of 0.15.
Figure 4.
We randomly picked 10 times of 3, 6, 9, 12, 15, 17 genomes from the remaining 18 genomes and computed the AUC scores of predicting Ecol. Error bars are representing 90% confidence intervals on the estimates of the means.
Figure 5.
AUC scores from cross-organism Geptop prediction.
The range of AUC is from 0.5 to 1.
Table 2.
Cross-organism test accuracies of the Geptop.
Figure 6.
The distribution of predicted essential gene numbers.
Figure 7.
AUC scores from cross-organism tests using the integrative compositional information predictor.
The range of AUC is from 0.5 to 1.
Table 3.
Comparison of features between the experimental and the Geptop groups.
Figure 8.
The connectivity distributions of four classes of genes in three bacteria with AUC lower than 0.7.
(a) H. influenza; (b) H. pylori; (c) V. cholera.