Skip to main content
. 2015 Sep 28;10(9):e0138913. doi: 10.1371/journal.pone.0138913

Fig 3. Performance and accuracy evaluation (%) of databases.

Fig 3

3A. Graphical representation of databases performance has been shown in percentages. BioGPS database revealed 97%, VeryGene database 92%, TiSGeD database 76%, TiGER database 66%, UniGene database 37%, C-It database 5% and the HPA unveiling 45% performance for the identification of liver-specific protein biomarkers. Performance % was calculated by dividing number of proteins identified by each database to total number of proteins that passed the filtering criteria. 3B. Graphical representation of accuracy of the initial protein identifications with HPA database showing the highest accuracy of 25%, VeryGene database showing 8% accuracy, TiSGeD database showing 15%, TiGER database showing 8%, UniGene database showing 19%, C-It showing 2% and BioGPS database showing 20% accuracy. The accuracy was calculated by dividing number of proteins that had passed the filtering criteria by each database to the total number of proteins each database initially identified.