The Recognition Of Five Ligand Binding Residues In Proteins

Interaction between proteins and ligands is the main way to carry out proteins’ function. So the recognition of ligand binding residues plays a important role in understanding proteins’ function, and it is very useful for drug design and development.ATP, ADP, GTP, GDP and NAD are very significant in biological cells. For example, ATP is a high-energy phosphate compound, it can hydrolyze and release much energy by hydrolyzing phosphate groups. ADP is the outcome. There is ADP in high density particles of platelet cells to expedite the platelet coagulation process. So Accurate recognition of ligand binding residues is much important. It is difficult to predict ligand binding residues considered time and money by the experimental means because of a large number of proteins and species. The theoretical calculation is useful and efficient way to achieve the goal.Based on sequence information, five ligand binding residues were recognized in proteins, our major works are as follows:(1) We built the ATP, ADP, GTP, GDP and NAD five ligand binding residues benchmark datasets, which included binding residues 3838,4865,1316,1701 and 3579, respectively,sequence identity is below 30% and resolution is better than 3?. Overlapping segments were generated by "sliding window". On the basis of calculation results of different window sizes, we determined the best window size as 17.(2) According to the biological character of five ligand binding residues, we establish WEBLOGO and analysis the differences between binding segments and non-binding segments.We study amino acid composition, amino acid conservative information,hydrophobicity, hydrophilicity polarity characteristics, second structure information,second structure conservative information and surface accessibility information. According to the result,we choose appropriate parameters to predict ligand binding residues.(3) Based on amino acid composition and the amino acid conservative information, we predicted the five kinds of ligand binding residues in proteins, which are characterized by the increment of diversity algorithm and matrix scoring value algorithm.The results are not satisfactory.(4)After that, while with increment of diversity values, matrix scoring values, auto covariance values of physicochemical property, second structure information and surface Accessibility information as the parameters for support vector machine, the overall prediction Accuracy and MCC of ATP, ADP, GTP, GDP and NAD achieved 77.4%, 0.549;71.2%, 0.414; 82.1%, 0.643; 82.9%, 0.659 and 85.3%, 0.702 by 5-fold cross-validation,respectively.