In this study, we present a method for the prediction of physiochemical properties of catalytic sites residues using a suitable Artificial Neural Networking (ANN) Feed Forward Backpropagation algorithm coupled with a set of structural proteins with the properties of their amino acid residues. The method has been applied to a set of 100 structural proteins from the Protein Data Bank (PDB) having a ligand at their active site. Using Ligplot program for searching of active site residues and Surface racer for identifying the non active site moieties, the identified amino acid residues were classified in 15 different categories based on their physiochemical properties. After classification of active and non active site amino acids, their properties were converted into machine language. Furthermore, we created Neural Network Using Matlab software and generated algorithm for training and testing of data. Thereafter, analysis of results showed that 95% of active site’s physiochemical properties were correctly predicted. It is hoped that this work would help in determining the surface topographic properties for ligand binding sites residues in protein. The computational outcome would be helpful in ligand designing, molecular docking, de novo drug designing and structural identification and functional sites Comparisons.
