The Identification of Metal Ion Ligand-Binding Residues by Adding the Reclassified Relative Solvent Accessibility

Abstract

Many proteins realize their special functions by binding with specific metal ion ligands during a cell's life cycle. The ability to correctly identify metal ion ligand-binding residues is valuable for the human health and the design of molecular drug. Precisely identifying these residues, however, remains challenging work. We have presented an improved computational approach for predicting the binding residues of 10 metal ion ligands (Zn^2+, Cu²⁺, Fe²⁺, Fe³⁺, Co²⁺, Ca²⁺, Mg²⁺, Mn²⁺, Na⁺, and K⁺) by adding reclassified relative solvent accessibility (RSA). The best accuracy of fivefold cross-validation was higher than 77.9%, which was about 16% higher than the previous result on the same dataset. It was found that different reclassification of the RSA information can make different contributions to the identification of specific ligand binding residues. Our study has provided an additional understanding of the effect of the RSA on the identification of metal ion ligand binding residues.

Keywords: binding residues; metal ion ligand; position weight matrix; relative solvent accessibility; secondary structure.

FIGURE 1

Flowchart of the method for the identification of metal ion ligand-binding residues.

FIGURE 2

The statistical distribution of relative solvent accessibility in positive and negative set for K⁺ ligand. Note: the abscissa axis is the values of the relative solvent accessibility, and the ordinate is the number of amino acids corresponding to each predicted value. The solid red line represents the positive set, and the dotted blue line represents the negative set.

FIGURE 3

The feature importance of Zn²⁺ ligand indicated by MeanDecreaseAccuracy value (A) and MeanDecreaseGini value (B) from Random Forest. Note: the larger the MeanDecreaseAccuracy and MeanDecreaseGini values, the higher the importance of the feature parameters. WA1-WA18 is the features of amino acid, QS1-QS18 is the features of hydrophobic, DH1-DH18 is the features of charge, SS1-SS18 is the features of secondary structure, and SA1-SA18 is the features of relative solvent accessibility.

FIGURE 4

The comparison of prediction performances between several machine learning methods based on the same features by using five fold cross-validation test.