|
Deep-WET: a deep learning-based approach for accurate prediction of DNA-binding proteins using word embedding techniques To use the server, users first need to input the protein sequence into the input box. After submission, the server will evaluate the protein sequence and check the format for processing. After completing the submitted predication task, the result will be displayed on a separate page. Users will be able to see the result part, where users can see the prediction probability of the protein. If only one protein sequence is inputted, sometime it takes time to show the prediction probabilities results. Here, the result depends on the length of the protein and the number of protein sequences. When a user gives large number of protein sequence as the input, the Deep-WET prediction will take much time. The comparatively long computational time derives from the fact that Deep-WET need to perform word embedding techniques and the process of DBPs prediction to obtain discriminative features and use the best parameters for CNN to predict whether the inputted protein sequence is DBPs or not. Moreover, for the limitation of computational resource, we strongly recommended that user need to input less than 10 protein sequences at a time. |