Predicting the protein structure using random forest approach

Predicting the secondary structure of proteins is a challenging task. A large variety approaches exist that include observation using equipment’s and theoretical evaluation, in which the optimal structure is determined. The secondary structure determines 3D tertiary structure of protein, on which features and functionalities of protein depend. This paper use classification technique, Random Forest to build a model which is able to determine structure of unknown proteins. The dataset included the amide frequencies of proteins whose structure is known. Machine learning model is developed that can predict the structure of protein that still need to be exploited. The accuracy of the model is determined using ROC curve. The results confirm the performance of the model constructed using amides dataset.