Molecular Recognition and Feature Extraction System

This chapter is based on the prediction of MoRF regions within the intrinsically disordered protein sequence. Disordered proteins have molecular recognition regions (MoRF) making them highly attractive to bind with protein pairs. Thus, as they combine with other protein pairs, they undergo disorder-to-order transition making them essential for various biological functions. Therefore, the project is tasked to obtain structural information of the disordered protein sequence and perform machine learning techniques to predict the MoRF regions in disordered protein sequences. The proposed method for the project will focus on programming and simulation analysis using the MATLAB software for which structural information will be extracted from the disordered protein sequences. Using these sequences, the project is aimed to perform training and testing implementation. Two test methods are used to evaluate the performance of the trained SVM models. Analysis has shown that the cross-validation test method outperforms the independent test method.