Protein Disorder Prediction Using Machine Learning Techniques

This work attempts to develop a computationally fast protein disorder prediction model that has a high sensitivity and stable MCC (Matthews Correlation Coefficient) score, when compared to similar predictors. Further, this work focuses on these goals to ensure a very low number of false negative predictions by the presented model. However, with this focus on sensitivity, the model may produce an increased amount of false positive predictions. For that, it is important to monitor the MCC score and make sure to keep it relatively high as well. Accordingly, this confirms the efficiency of the presented model. The obtained results recommend the use of model developed for disordered protein prediction.