Figure 5
From: Integrating phenotype and gene expression data for predicting gene function
Total number of correct positive predictions. As the cutoff used in the prediction algorithm is increased, all of the datasets make fewer correct positive predictions. The integrated data set makes nearly as many correct predictions as the gene expression data set, and they both make many more predictions than the phenotype data set. This confirms that the phenotype data set is not as capable as the other data sets in predicting new annotations. The generalized predictions always result in more true positives. Figure 5a shows the results for the generalized predictions. Figure 5b shows the results for the exact predictions.