Details for misclassified proteins in Dataset 3 using HybSVM.

Details for the misclassified proteins in Dataset 3 based on using the HybSVM method including: actual class (multi, singlish), gene name, and number of interfaces as predicted by Kim et al. [16] are shown. For each misclassified protein, information about the protein with the highest homology based on the nearest BLAST hit is also shown. This information includes: gene name, number of interfaces as predicted by Kim et al. [16], e-value of the BLAST results between the two proteins, and the difference between the number of predicted interfaces for the misclassified protein and its nearest BLAST hit.