Predicts mycobacterial membrane proteins and their types. MycoMemSVM is an online server and a binomial distribution-based feature selection technique able to select over-represented tripeptides. In the development of this method, the binomial distribution was used to pick out the over-represented tripeptides and the support vector machine (SVM) was used to perform prediction. This server can be helpful for the vast majority of experimental scientists who focus on mycobacterium and antimicrobial drugs.
Recognizes mycobacterial membrane protein and their classes. Unb-DPC uses oversampling technique Synthetic Minority Oversampling Technique (SMOTE) to remove biasness among different type of member proteins. It utilizes dipeptide compositions to extracted features from the unbiased data. This tool is able to avoid biasness among different classes and preserves protein sequence structure information simultaneously.