Prediction of DNA Interacting Residues

DNA residue interaction is one of the most important interaction in the biological system. It is used for describing the working of various biological processes. In this model we tried to predict the DNA interacting residue by training the model using dataset having around 38000 data elements and testing it using dataset having around 2000 data elements. The prediction is done using various supervised machine learning algorithms namely - the decision tree, support vector machine, logistic regression and linear discriminant analysis and subsequently the different accuracies and F-scores are noted. The accuracy of decision tree is 50.6%, support vector machine is 48.9%, logistic regression is 48.6% and linear discriminant analysis is 48.9%. The best algorithm - decision tree is selected based on the accuracy and F-score and can be used for further requirements in the fields such as hospitals, healthcare centres, testing labs, research labs, genetic study. The prediction is done in a noble and effective way and efforts have been made to make the paper unique, usable and highly successful.