Machine Learning Classifiers based on Predicting Membrane Protein using Decision Tree and Random Forest

Bioinformatics protein data was used for drug design, finding diseases, a gene for a living organism, immune relation, etc. Analysis of protein data is hard to see its full range functions and fulfill the biological queries. Protein primary information are available in amino acid sequence datasets. Therefore, based on the primary biological data, it is highly sensible to construct an efficient predictor for the identification of protein function and their types. In this study, imbalance and excessive datasets are often managed well by machine learning classifiers. In this study, machine learning classifiers build on the features of membrane cell sequence and some of their essential functions depend on PseAAC (Pseudo Amino Acid Composition) descriptors. In this study proposed Prediction model based on decision tree classifiers such as Information gain, Gini-Index, and Random Forest and performance analyzed based on their resultant accuracy among the classifiers, the random forest was performed well, were the accuracy of 91.67%.