Figure 6
From: Integrating phenotype and gene expression data for predicting gene function
Precision. As the cutoff used in the prediction algorithm is increased, the precision of all of the data sets increases. Precision is defined as (tp)/(tp + fp) [16]. Combined with Figures 4 and 5, this indicates that, while fewer false positives are predicted as cutoff is increased, fewer true positives are also predicted. This is especially true in the case of the phenotype data set, which resulted in far fewer predictions than the other data sets. The integrated data set does outperform the other data sets. The generalized predictions result in a better precision than the exact predictions. Figure 6a shows the results for the generalized predictions. Figure 6b shows the results for the exact predictions.