Welcome to ThpPred

ThpPred is a prediction tool developed to screen therapeutic proteins vs non-therapeutic proteins. Users can consult our database THPdb2, which is particularly designed to store all information on therapeutic protein. Relavant information related to the therapeutic proteins used to construct this prediction server are available here (See ThPdb2). This method is developed using FDA approved therapeutic peptides. We have implemented various machine learning algorithms. The best performing model have been applied in the webserver. In ThpPred, we have developed an ensemble method combining motif search analysis with Random Forest models to assess if a protein is therapeutic or not. We used a dataset of 356 therapeutic and 356 non-therapeutic proteins to train our models. This server contains four major modules such as; i) Prediction , ii) Design, iii) Motif Scan, and iv) Download