Skip to main content

Advertisement

Figure 2 | BMC Bioinformatics

Figure 2

From: Towards the prediction of essential genes by integration of network topology, cellular localization and biological process information

Figure 2

ROC curves and AUC values for the classifiers trained on balanced datasets with individual or grouped network topological features. ROC curves and AUC values of classifiers trained on balanced dataset 9 (see Figure 1) with one or groups of network topological features as learning attributes as follows: "net": all network topological features as learning attributes; "ppi", "inbetppi", "inbet", "c", "cent", "regin", "metin", "regout". "metout", "inbetmet", "ident" and "inbetreg": datasets with only one of the following network topological features as learning attribute: number of protein physical interactions (ppi), betweenness centrality for the protein physical interactions (inbetppi), betweenness centrality for all types of interactions (inbet), clustering coefficient (c), closeness centrality (cent, number of regulating transcription factor (regin), number of reactants participating in a metabolic reaction catalyzed by the enzyme encoded by the gene (metin), number of genes regulated by the transcription factor encoded by the gene (regout), number of products generated in a metabolic reaction catalyzed by the enzyme encoded by the gene (metout), betweenness centrality for the metabolic interactions (inbetmet), number of genes with identical topological features (ident) and betweenness centrality for the transcriptional regulation interactions (inbetreg). "only _ppi_metabolic_features" and "only_ppi_reg_features": datasets containing protein physical interactions-related features (ppi and inbetppi) and, respectively, metabolic (met, metin, metout and inbetmet) and transcriptional regulatory interactions-related features (reg, regin, regout and inbetreg). "only_ppi_other_features": dataset containing protein physical interactions-related features (ppi and inbetppi) and c, ident, cent and inbet. "only_ppi_inbetppi": dataset containing only the indicated network topological features as learning attributes. For more details on network topological features, see Additional file 1.

Back to article page