Figure 1.
The relationship among the HCV proteins, the target genes and the between genes.
The yellow node (V1, V2) are HCV proteins. The target genes (blue nodes, T1, T2 and T3) were the human target genes of HCV proteins. The between genes (red nodes, B1, B2 and B3) were human genes that were on the shortest paths between target genes in protein interaction network. The grey nodes were other human proteins that were neither target genes, nor between genes.
Figure 2.
The IFS curves of target genes and between genes.
In the IFS curve, the X-axis is the number of genes used for classification, and the Y-axis is the prediction accuracies of nearest neighbor algorithm evaluated by the jackknife test (or LOOCV). The blue curve is the IFS curve of target genes and the highest accuracy was 0.944 with 155 genes. The red curve is the IFS curve of between genes and the highest accuracy was 0.960 with 162 genes.
Figure 3.
The number of selected target genes of each HCV protein.
The HCV proteins NS3, NS5A and CORE have the largest numbers of target genes in the optimal set of the 155 selected target genes.
Figure 4.
The number of selected between genes for each of the HCV protein pairs.
The following pairs have more than 80 selected between genes: NS3_NS5A, CORE_NS3, F_NS3, E2_NS3, NS3_NS5B, CORE_NS5A and E1_NS3.
Table 1.
The enrichment of the 155 selected target genes and the 162 selected between genes on OncoDB.HCC genes.
Table 2.
The KEGG enrichment of the top five target genes (EFEMP1, JAG2, TACSTD2, STAT3 and STAT1) and the top five between genes (PDIA3, LCP2, IL23A, SCAMP3, ISG15).