Investigating the Performance of Naive- Bayes Classifiers and K- Nearest Neighbor Classifiers

Abstract

Identifying binding motifs or critical sequence segments is a key step toward understanding the mechanism of interactions between hepatitis C virus (HCV) and host cell proteins, such as human proteins. In the present study, we found a pair of binding motifs, G185-L234 in HCV NS5A proteins and L-X(3,5)-E-[AEGNQST] in human proteins, which are considered to be a key determinant for the interactions between HCV NS5A and human proteins. The binding motif of human proteins often forms a full helix (H) or an extended strand-loop (EL) structure. We also found 3 highly conserved sequence segments in HCV NS5A, G185-D196, 7^216-^239, and S 414-8437, which are in good agreement with the experimental results of previous studies. These results are expected to prove helpful in the design of high-affinity molecules that target the binding sites of HCV NS5A and human proteins.